Thesis
The modification of perception of emotions by transcranial magnetic
stimulation
ROCHAS, Vincent
Abstract
L'objectif de ce travail de thèse est d'étudier le potentiel de la stimulation magnétique
transcrânienne (TMS) et de ses effets à court terme sur la perception des émotions. La
modulation au niveau comportemental de la perception des émotions est étudiée à travers
deux travaux indépendants portant sur la perception des émotions dans les expressions
faciales ou la détection des émotions dans les mots. Dans les deux cas, les cibles et le
moment pour l'application de la TMS ont été guidés par des résultats d'électrophysiologie. La
perturbation de la pré-SMA conduit à la diminution de la reconnaissance de l'expression
faciale de joie alors que la perturbation de la jonction temporopariétale gauche ou droite
entraîne le ralentissement de la détection des mots émotionnels en particulier dans le champ
visuel gauche. Ces résultats confirment une bonne correspondance entre les observations
faites en électrophysiologie et les effets dus à la perturbation par TMS.
Reference
ROCHAS, Vincent. The modification of perception of emotions by transcranial
magnetic stimulation. Thèse de doctorat : Univ. Genève et Lausanne, 2014, no. Neur. 131
URN : urn:nbn:ch:unige-369399
DOI : 10.13097/archive-ouverte/unige:36939
Available at:
http://archive-ouverte.unige.ch/unige:36939
Disclaimer: layout of this document may differ from the published version.
UNIVERSITE DE GENEVE
FACULTE DES SCIENCES
Département de neurosciences fondamentales
Functional Brain Mapping Laboratory
Pr Christoph M Michel
___________________________________________________________________________
The Modification of Perception of Emotions
by Transcranial Magnetic Stimulation
THESE
présentée à la Faculté des sciences de l’Université de Genève
pour le grade de Docteur ès sciences, mention Neurosciences
par
Vincent ROCHAS
de
Pierrelatte (France)
Thèse n° 131
GENEVE
REPROMAIL - Atelier de reprographie à UniMail
2014
Publications
The present thesis work was based on the following articles:
Rochas V, Gelmini L, Krolak-Salmon P, Poulet E, Saoud M, Brunelin J, Bediou B. 2013. Disrupting
pre-SMA activity impairs facial happiness recognition: an event-related TMS study. Cereb
Cortex, 23(7): 1517-25.
Rochas V, Rihs TA, Rosenberg N, Landis T, Michel CM. 2014. Very early processing of emotional
words revealed in temporoparietal junctions of both hemispheres by EEG and TMS. Exp Brain
Res, 232(4): 1267–81.
Remerciements
Je remercie de manière générale mais de manière totale le Functional Brain Mapping Laboratory du
Prof. Christoph M. Michel de l’Université de Genève qui m’a offert le meilleur environnement
imaginable pour effectuer mon doctorat. C’était tellement bien que j’ai failli ne pas finir !
Je remercie tout particulièrement Christoph pour m’avoir si bien accueilli, humainement, et de s’être
toujours occupé de moi en me laissant (peut être trop) mener mes projets comme je l’entendais. Pour
son soutien, pour sa patience qui aura été grande, et pour les interactions scientifiques (et autres) que
nous avons pu avoir. Pour m’avoir formé et amélioré. Et bien sûr pour m’avoir transmis son savoirfaire propre à l’utilisation de l’EEG haute densité, le saint-graal de l’électrophysiologie.
Je remercie Thedi pour m’avoir confié son projet qu’il affectionne tant et que j’espère avoir au moins
en partie étudié correctement. Je remercie Mélanie pour sa gentillesse, son accueil, et sa bienveillance
à l’égard d’un petit nouveau que j’étais en arrivant. Je remercie Juliane pour m’avoir montré toutes les
petites feintes qui facilitent la vie du bon scientifique qu’on se doit d’être, et pour la bonne tenue du
labo. Je remercie Denis sans qui nous ne serions pas qui nous sommes. Pour les discussions
inattendues de fin de repas au CMU et aussi pour sa compassion à mon égard les durs jours d’hiver où
je m’obstine à venir en moto. Je remercie Laurent pour prendre tellement soin de notre peau et nous
préserver du moindre rayon du soleil (!), aussi pour son niveau de blagues acérées et absurdes qui nous
vaut bien des duels. Je remercie Cristina Berchio for her Chinese discourse and her absolutly
irrepressible good mood which is though full of stress! Je remercie évidement Maria pour m’avoir
toujours soutenu, pour son avis avisé sur les statistiques et pour être entière et avoir de vrais états
d’âme. Je remercie Miralena pour ces bons moments partagés et pour m’avoir montré comment mettre
carpette un belge au levé de coude ! Je remercie Charles (sans rancune) pour n’avoir jamais évoqués
un sujet en rapport avec le travail tout en étant au travail. D’ailleurs ça n’a rien à voir mais on doit
toujours aller sur le Salève. Je remercie Holger pour son monstre bon esp’ et pour ne jamais se prendre
la tête. Aussi pour avoir donner forme à l’objet le plus incongru du CMU. A Lore pour sa bonne
humeur. Je remercie Gijs pour son fair-play et son esprit éclairé. Je remercie Jennifer et Samir pour
avoir rendu tolérables ces centaines d’heures passées dans le BBL. Je remercie Seb et Bruno pour
toutes ces coupes.
Je remercie tout particulièrement Alexis chez qui j’ai trouvé un peu un autre grand frère et avec qui
j’ai beaucoup partagé (et je compte bien continuer) et surtout énormément rigolé (et ça aussi je compte
bien continuer). Aussi pour son regard aigu sur des questions d’importance et son regard aigu sur des
questions qui importent peu. Pour tout le reste aussi. Je remercie Robert avec qui c’était bien cool de
partager le bureau et pour les bons moments en dehors du bureau. Pour son amitié et because he
maintained alive my few remaining skills in English (could you imagine how it would be without
him?). Bien sûr aussi pour sa collaboration à l’excellente recette du Fantomas. Je remercie Grig pour
avoir prouvé au monde qu’on pouvait aimer le Dubstep et la Trap en étant (presque) normal !
Je remercie ceux qui sont partis en cours de route mais qui compte quand même, Laura, Nadia et
Verena. Aussi Karsten avec qui je n’ai pourtant jamais pu aller au Palais Masquotte… Je remercie
Arnaud et Yann (qui ne sont en fait pas vraiment partis, ah bin si en fait Yann vient de partir) pour
avoir été les légendes qu’ils sont.
Je remercie enfin Tonia sans qui je n’aurais peut être à remercier personne car je me serais trouver
tellement perdu que j’aurais peut être simplement tout arrêté. Sans Tonia il n’y aurait pas grand-chose.
Merci pour cette agréable collaboration et cette interaction quotidienne. Merci pour ces milliers de
précieux conseils. Merci de m’avoir bien guidé dans mon travail et en dehors, de m’avoir formé. Merci.
Je remercie ma fraise, ma muse, ma meuf, Marie. Celle qui m’inspire au quotidien. Pour son soutien
sans faille, dans l’ombre, mon Bernardo (  ) que je ne remercierai jamais assez même si je compte
bien m’y atteler encore très très très longtemps. Tu n’en as pas fini avec moi ! Mi amor…
Je remercie ma famille et mes parents, parce que ça se fait et surtout parce qu’ils m’ont toujours
encouragé et ne m’ont jamais posé aucune limite. Jamais. Je remercie mon frère parce qu’il est
vraiment comme un frère pour moi. Mais vraiment !
Je remercie Ben & Jé qui pourrait être une marque de biscuits apéritifs mais qui ont surtout formé mon
jeune esprit avant que tout cela ne commence vraiment et qui m’ont donné goût à la recherche. Mes
premiers mentors. A feû l’EA4166 de Lyon.
Je remercie également les jurés pour le temps et les conseils et améliorations qu’ils ont pu fournir à ce
travail de thèse.
Je m’arrêterais là car cela commence à ressembler à des adieux, mais j’oublie obligatoirement des
gens et je m’en excuse.
Table of Contents
RÉSUMÉ / ABSTRACT..........................................................................................................................1
I. INTRODUCTION ...............................................................................................................................3
II. EMOTION PERCEPTION ..................................................................................................................4
1. Emotion and brain? ....................................................................................................................4
2. Emotional classes .......................................................................................................................5
3. Emotion in cognition ..................................................................................................................7
3.1. Emotion as a function ...........................................................................................................7
3.2. Emotion and cognitive impairment .......................................................................................8
4. Emotion modalities ....................................................................................................................9
4.1. The different emotional modalities.......................................................................................9
4.2. Focus on two modalities, faces and words .......................................................................... 10
5. Neural structures in emotion and potential targets ................................................................. 11
5.1. Emotion structures and organisation .................................................................................. 11
5.2. Neural treatment of facial emotion expressions ................................................................. 14
5.3. Neural treatment of emotional language material .............................................................. 16
5.4. Right hemisphere in reading of emotional words ................................................................ 18
III. TRANSCRANIAL MAGNETIC STIMULATION AND APPLICATIONS .................................................. 19
1. History of TMS: from bioelectricity to modern TMS ................................................................. 19
2. Principles of TMS ...................................................................................................................... 21
2.1 Mechanisms of stimulation: from electric current to cell depolarisation .............................. 21
2.2 Under the coil: distribution of currents ................................................................................ 22
2.3 Different current waveforms and their implications............................................................. 24
2.4 Different coil types and their incidences .............................................................................. 27
2.5 Coil handling........................................................................................................................ 28
3. Protocols of stimulation ........................................................................................................... 30
3.1. Single pulse TMS................................................................................................................. 30
3.2. Repetitive TMS (rTMS) ........................................................................................................ 31
3.3. Patterned and Theta burst stimulation (TBS) ...................................................................... 31
3.4. Paired TMS ......................................................................................................................... 33
4. Interests of the technique ........................................................................................................ 33
4.1. Use of TMS as a research tool ............................................................................................. 33
4.1.1. Measuring neural excitability ....................................................................................... 34
4.1.2. Modification of excitability and long term effects ........................................................ 36
4.1.3. Measuring indirect effect on behaviour ....................................................................... 37
4.1.4. Performance enhancement ......................................................................................... 39
4.1.5. Tracking of the TMS impact with brain imagery ........................................................... 41
4.2. Use of TMS as a clinical tool ................................................................................................ 44
5. Controls .................................................................................................................................... 46
5.1 Control TMS ........................................................................................................................ 46
5.2 Control conditions ............................................................................................................... 48
6. Localisation and neuronavigation ............................................................................................. 48
6.1. Target definition based on existing literature...................................................................... 49
6.2. Target definition from individual imaging ........................................................................... 49
6.3. Positioning methods ........................................................................................................... 50
7. Safety ....................................................................................................................................... 51
7.1. Direct effects due to the TMS magnetic field generation .................................................... 51
7.2. Indirect effects of TMS on biological systems...................................................................... 53
IV. THE USE OF ELECTRICAL NEURONAL SIGNALS TO GUIDE TMS ..................................................... 54
1. Intracranial electrodes for target localisation ........................................................................... 54
1.1. Recording ........................................................................................................................... 55
1.2. Deep brain stimulation ....................................................................................................... 55
1.3. Target localisation. ............................................................................................................. 55
2. EEG for target localisation ........................................................................................................ 56
2.1. Evoked potential and waveform analysis ............................................................................ 56
2.2. Topographic analysis .......................................................................................................... 57
2.3. Source localisation.............................................................................................................. 58
V. HYPOTHESIS ................................................................................................................................. 59
VI. DESCRIPTION OF THE ARTICLES ................................................................................................... 59
1. Disturbance of facial expression recognition by TMS ............................................................... 59
1.1. Summary of the results....................................................................................................... 59
1.2. Contributions to this work .................................................................................................. 60
2. Lexical detection of emotional words ....................................................................................... 60
2.1. Summary of the results....................................................................................................... 60
2.2. Contributions to this work .................................................................................................. 60
3. Contribution to additional works ............................................................................................. 61
VII. DISCUSSION ............................................................................................................................... 62
1. Foreword .................................................................................................................................. 62
2. Behavioural measure and event-related TMS paradigm........................................................... 63
3. Potential of electrophysiology for TMS target definition.......................................................... 66
4. The role of the premotor cortex in emotions ........................................................................... 69
4.1. Motor areas for emotional recognition ............................................................................... 69
4.2. Alteration of the mimicry system ........................................................................................ 74
4.3. A prefrontal question ......................................................................................................... 76
4.4. A functional model around the target ................................................................................. 77
4.5. Implications ........................................................................................................................ 78
5. The right hemisphere an early reader of emotions................................................................... 79
5.1. Processing of emotions in the right hemisphere ................................................................. 79
5.2. Pre-processor of written material in the right hemisphere .................................................. 81
5.3. A functional model around the targets ............................................................................... 82
VIII. CONCLUSION ............................................................................................................................. 84
IX. REFERENCES ................................................................................................................................ 84
X. ARTICLES .................................................................................................................................... 107
1. The first study using TMS: on facial expression recognition ................................................... 107
2. The second study using EEG and TMS: on lexical detection of emotional words .................... 121
RÉSUMÉ / ABSTRACT
L'objectif de ce travail de thèse est d'étudier le potentiel de la stimulation magnétique
transcrânienne (TMS) et de ses effets directs et à court terme sur la perception des émotions. La
modulation au niveau comportemental de la perception des émotions est étudiée à travers deux
travaux indépendants portant sur différents substrats émotionnels. Un travail explore la modification
de la perception des émotions dans les expressions faciales; le deuxième travail quant à lui explore la
détection des émotions dans les mots. Les deux travaux étudient l'implication de structures
inhabituelles ou non conventionnelles dans le traitement des émotions. La première aborde le
traitement émotionnel par le biais d’une aire du cortex prémoteur en possible relation avec un
phénomène de mimétisme impliquant des mécanismes « miroirs ». Le deuxième travail étudie le rôle
d’aires corticales de l'hémisphère droit dans le traitement précoce des mots écrits. Dans les deux cas,
les cibles et le moment pour l'application de la TMS ont été basés sur des résultats
d'électrophysiologie. Le premier travail se base sur des études existantes et utilisant des électrodes
intracrâniennes dans l’aire pré motrice supplémentaire (pré- SMA) de patients. Dans le second cas, la
TMS a été guidée par une expérience indépendante d'électroencéphalographie chez des participants
sains. Ce deuxième travaille utilisant la TMS évalue l'implication des jonctions temporoparietales
(TPJ) dans la détection de mots émotionnels. La TMS dans ces aires non traditionnelles des émotions
a démontré de possibles modifications d’aspects spécifiques de la perception des émotions. La
perturbation de la pré-SMA conduit à la diminution de la reconnaissance de l'expression faciale de
joie alors que la perturbation de la TPJ gauche ou droite entraîne le ralentissement de la détection
des mots émotionnels en particulier dans le champ visuel gauche. L’électrophysiologie a ainsi guidé la
pratique TMS, et ces résultats confirment une bonne correspondance entre les observations faites en
électrophysiologie et les effets dus à la perturbation par TMS.
The objective of this thesis work is to investigate the potential of the transcranial magnetic
stimulation (TMS) and its direct short-term effects on the modification of the perception of emotions.
The behavioural modulation of the perception of emotions is studied through two independent
works on different emotional materials. One work explores the modification of the perception of
emotions in face expressions. The second work explores the detection of emotions in written words.
The two works investigate the implication of non-conventional structures in the emotion processing.
The first work attacks emotion processing through an area of the premotor cortex in possible relation
with mimicry phenomenon implicating mirror mechanisms. The second work studies the role of the
1
right hemisphere cortical areas in the early processing of written word materials. In both cases, the
targets and the timing for the application of event-related TMS was based on electrophysiology
results. The first work was based on existing studies using intracranial electrodes in the pre
supplementary motor area (pre-SMA) of patients. In the second case the TMS was guided by an
electroencephalography investigation in independent healthy participants. This second TMS
experiment assessed the implication of the temporoparietal junctions (TPJ) in the detection of
emotional words. Event-related TMS on these non-traditional areas in emotions demonstrated
capabilities in the modifications of specific aspects of the perception of emotions. The disruption of
the pre-SMA leads to the diminution of the recognition of facial happiness expressions when the
disruption of the left or right TPJ leads to the slowing of the detection of emotional words especially
in the left visual field. Electrophysiology was able to guide TMS practice, and these results confirmed
a good matching between the observations from electrophysiology and the effects due to the
disruption by TMS.
2
I. INTRODUCTION
Emotions are crucial and particularly developed features of human beings. Their contribution
in human behaviour is primordial and they constitute an important part of communication and
cognition. They are also in return responsible for many neuropsychological troubles. On a functional
level, emotions can be divided in three main parts that are not always easily dissociable: perception,
expression and introspection. Inducing modulation of emotion perception can be envisaged as a
mean to modify their treatment in healthy individuals or to restore certain altered functions in
pathologic situations. Obviously, the understanding of the mechanisms of modulation is also a
manner to understand the perception of emotions and its malleability.
The brain is considered as the main actor to process and create emotions even if a more integrated
view considers the relations between the brain and the rest of the body as crucial mechanisms of
emotional processing. Emotions are in link with different other cognitive functions and are conveyed
in a large cerebral network. However a bodily integration of emotion signals by the motor and
premotor areas is still debated. Moreover the contribution of the right hemisphere is also not clear
especially in the language modality dominated by the left hemisphere.
Transcranial magnetic stimulation (TMS) is a non invasive technique of stimulation of the cerebral
cortex that can induce transient lesions. As a tool of functional modification of the brain, it
constitutes a good candidate for the modification of emotion perception.
In this thesis, we will use TMS to investigate the cerebral processes involved in different aspects of
emotion perception. Electromagnetic techniques will guide our stimulation procedure by defining the
time as well as the location to be targeted by event-related TMS. In Introduction we will first discuss
the mechanisms of emotion processing and focus on the emotional stimuli that will be particularly
studied in this thesis: faces and words. It will then focus on TMS and describe the method and its
potential applications in detail. In another chapter, it will briefly describe the methods that allow for
the description of the brain processes in time as well as in space with electrophysiology and evoked
potentials. The thesis will include two studies, one study using TMS based on previous intracranial
results and one study using evoked potential from high-density electroencephalography in order to
guide the timing and localisation of the TMS. The results from these two studies will be discussed in
view of modulation of emotional processing.
3
II. EMOTION PERCEPTION
The substantive “emotion” is commonly defined as a feeling or an affective state usually
accompanied by certain physiological markers. In this sense, it is a psycho-physiologic experience of
the mental and body state in response to internal and external signals – i.e. endogenous biochemical
molecules and perceived information from the environment. Conforming to the scientific or the
common language, the term emotion could refer to a psychological state, a scene or a music for
instance. In concrete situations, it is often nuanced as a fluctuating cocktail of several affective states
oscillating between extremes. In fact, its etymology refers to the notion of movement (motio in Latin),
that correspond to its volatility – as opposed to mood that stays for a while – and to its propensity to
be linked to and to generate movements or characteristic behaviours. Note that emotions are often
defined for humans as it is the case in the present work but are not necessarily restricted to humans.
Finally, the multiplicity of aspects of a definition for the term tells us that emotion is a tricky and
wide concept.
1. Emotion and brain?
Through history of civilisations and the history of science, the place responsible for emotion
has been relayed to different structures of the body, sometimes in the heart, in the body fluids (also
named humours), in the brain or even outside of any physical feature in the soul. For instance Greek
philosopher Aristotle (384-322 BC) defined pain as an emotion and its seat was considered to be in
the heart. Hence these postulates are counter classical views on emotions in modern neuroscience
since the emotions are considered to be processed in the brain and, in addition pain – at least in its
somatic compound – does not enter in a straightforward and classical emotional categorisation. The
question of interactions between the brain and the heart, and more exactly between the brain and
the body as a whole, was strongly debated in the early 20th century. This debate originated from the
theory of emotion of psychologists William James and Carl Lange in 1884 who stated that the body
and its automatic reactions are the basis and even the cause of emotion. This assumption was then
refuted by works of the physiologist Walter Cannon and Philip Bard who proposed an alternative
theory – the Cannon-Bard theory (Cannon, 1927) – in favour of a cerebral genesis of emotion.
However the complex and antagonist duality opposing heart and brain as a centre for emotion and
beyond for the control of the individual, remained existing during the middle of 20th century until the
advent of modern neuroscience – first with neuropsychology cases and then with imaging studies
which clearly identified the brain as the seat of emotion. However the interaction between brain and
body is important in the study of emotion as well as in a variety of other functions and behaviours.
4
The brain, often considered as the only control on thought and behaviour, interacts continuously
with the whole organism and integrates its messages and state changes. It is particularly true in
emotion processing with numerous external markers – e.g. hormones – that are produced by the
body in response to emotions and may interact with cognitive emotion processing in return.
2. Emotional classes
The classification of emotions is also a subject of vigorous debates. Some authors favour a
categorical approach postulating a finite number of primary or fundamental emotions. There is
currently no consensus on how to define these primary emotions and their types and numbers.
However, the classification the most common and accepted and the most used in research results
from the works and theory of the American psychologist Paul Ekman. Ekman applied the principles of
evolution of Darwin (Darwin, 1872) to the domain of emotions. According to his theory, emotions
relay on universal mechanisms probably because of their primordial role in the individual survival in
the group and the society. He postulated that emotion is perceived and produced the same way all
over the world independently of ages, gender, culture, or ethnicity (Ekman and Friesen, 1971). Hence
the processing of emotion shares a common basis in any individual with common neural structures
and acts as a reflex with common pattern of actions. This theory stated the model of Basic Emotions
(Ekman, 1973; see figure 1). According to this model, one can distinguish six basic emotions that are
happiness, sadness, anger, fear, disgust and surprise. Importantly these six first basic emotions were
based on facial muscle movement. However emotions in humans are not limited to six states or not
even to mixes of these six, but secondary emotions have been later described – also by Ekman
himself – corresponding to social, cultural, or moral criterion and finer feelings. One can mention
shame, contempt, guiltiness, amusement, contentment, embarrassment, excitement, guilt, pride,
relief, or satisfaction. Despite the really well established place for Ekman’s work in psychological and
neuroscience studies, the supposed universal and innate nature of the recognition and of the
production of emotions is controversial and judged as a reduced insight on emotion perception
(Barrett, 2006). Notably the use of forced choices could represent a critical limitation for
investigation of emotional recognition.
5
Figure 1: Example pictures extracted from the original Ekman and Friesen, 1979. From left to right: Neutral, happy, sad
and fear facial expressions are represented in one unique identity.
Wrongly opposed to the categorical classification, a dimensional vision of emotions has been
developed in order to classify emotions. The dimensional approach can integrate various dimensions
and axes. There is, however, no consensus on the number or the types of these dimensions. Also the
different dimensions are not necessarily exclusive. One can first dismantle emotions on dimensions
of arousal – i.e. the intensity of the feeling or the amplitude of emotional reaction – and valence –
e.g. positive for happiness or negative for sadness. Interestingly, the arousal dimension defines also
the neutral affective state at the lower end of this dimension with low arousal. According to the
valence dimension authors proposed the idea of a cerebral lateralisation in processing of emotions
according to their valence (Wager et al., 2003; details in section 5.1. Emotion structures and
organisation). The affective motivational aspect of emotions constitutes a third dimension. Thus one
can consider the different emotions according to their motivational effect; in other words in defining
their tendency to motivate approach and action or conversely to induce withdrawal behaviour and
rest. This last type of classification is easily transposable between the species suggesting a
phylogenetic link between species for this homologous trait. Any animal shows a motivated
behaviour of approach toward attractive stimuli – as a reward – and withdrawal of aversive stimuli –
as a punishment. This behaviour is found in the domain of human emotions as well which are prone
to increase or decrease the tendency to action or influence the decision making. Also particularly
important when considered in social interactions, emotions could lead to pro-social behaviour or on
the contrary to avoidance of others. Based on this last dimension, psychologist and psychiatrist
Richard Davidson (Davidson, 1992; Davidson, 2004) exposed a model of separation of emotions in
two categories according to a propensity to approach on one side with happiness and anger, or a
propensity to withdrawal with fear, sadness and disgust on the other side. Surprise could be rather in
the second category. The neural substrates of approach and withdrawal emotions are also differing
in term of hemispheric specialisation.
6
The three main dimensions of emotional categorisation exposed above are evidently complementary
or can be completed with others. Cabanac for instance proposed a classification of emotions on four
axes with intensity that is completed with quality (how well it is perceived), hedonicity and timing of
the emotion (Cabanac, 2002). Despite important divergences, the different dimensions in
classification of emotions are yet not incompatibles; each of them answers different questions.
Classification of emotions is important in cognitive neuroscience, as it is often a primordial
step to define specific mechanisms and functionalities according to emotion categories. Note that
scientific research studies use these classifications in order to control external stimuli classes and
hence aim at defining only elicited emotions. The study of spontaneous affective state is trickier.
3. Emotion in cognition
3.1. Emotion as a function
The emotion thematic aroused a new enthusiasm in scientific research thirty years ago with
the study of patients with cerebral lesions, electrical stimulation, electrical recordings and then more
widely by functional imaging that allowed to evidence the implication of a specific neural substrate in
specific emotional functions. Researchers in neuroscience, as Antonio Damasio in humans or Joseph
Ledoux in animals, raised a sharp interest for what was considered erstwhile as a somatic
phenomenon rather than neurological.
The vast majority of authors converge on the existence of three main emotional compounds that are
motor (e.g. facial, vocal or gesture expression), physiologic (e.g. autonomic, hormonal and cerebral)
and subjective (e.g. conscious feeling). Other compounds can be added as the tendencies to action or
motivational effect and the influence of emotions in the cognition (e.g. memory or decision making).
Emotion acts in potential interaction with other functions. On a cognitive level, emotion has been
defined as a proper cerebral function and eventually divided in sub-functions. It comprises
perception, integration, and production. Emotional perception could concern exogenous signals as
external stimuli from the environment (e.g. emotional words reading) and signals directly expressed
by other individuals (e.g. facial expression recognition) or endogenous signals (e.g. interoception of
autonomic reactions). The integrative part of emotion could then be involved both in perception with
the categorisation and identification of the endogenous and exogenous emotional signals and in
production with the inverse transduction of the internal state to a reaction. There is an integration of
the diverse compounds within cognitive processes (e.g. comparisons of current emotional signals
with memorised concepts) and a regulation of the emotional state (e.g. by mechanisms of
rationalisation and reinterpretation). The integration of emotions might also engage an introspective
7
process. The production part is the expression of emotion signals by various modalities that are body
movement, facial expression, vocal or verbal language in humans. The expression of emotion is a
more or less conscious process in which individual translates its own affective state into external
emotional signals. These diverse parts or compounds of the emotional function respond to different
needs and conditions. The concept of emotion is sometimes tricky to apprehend in research as its
different aspects appear linked. Moreover emotions in their multiple forms integrate the functioning
of the individual and interact with other functions. Neuroscience studies investigate the neural
mechanisms related to the various aspects of emotion and try to dismantle them when possible. The
interactive relation between these various aspects is also considered.
3.2. Emotion and cognitive impairment
The different aspects of emotions – i.e. perception, integration or expression – are crucial in
the correct functioning of any individual. They determine the correct communication and participate
in language processing. Emotions help the individual to judge his own situation and to take decisions.
Moreover they are the vector of unambiguous and rapid messages and constitute an advantageous
inter-individual relationship – e.g. in frightening situation. The deficit in one or several emotional
functions is a real handicap for the individual who will be in return not able to experience a proper
socialisation with adequate emission and/or reception of emotional signals and a fine integration of
the emotional sense. The cognitive regulation of emotional affect has been proposed as the basis of
many psychopathologies (Davidson, 2000). The perception of emotions also is thus variably lacking in
many pathologies. Schizophrenia patients show emotion recognition troubles (Bediou et al., 2007)
and modified brain responses to emotional stimuli (Sanjuan et al., 2007). In Alzheimer disease
patients and more generally but in a lesser extent in patients with mild cognitive impairment have
been found to have deficits in emotion recognition (Spoletini et al., 2008) and also in emotional
memory benefit (Kensinger et al., 2004). Deficit and bias in arousal and memory toward negative
emotions is also present in depressive patients (Atchley et al., 2007; Liu et al., 2012). The detection of
facial expression of disgust is impaired in OCT patients (Sprenglemeyer et al., 1997) and it would be
dependent on the severity of the troubles (Corcoran et al., 2008) or in Huntington disease patients
(Gray et al., 1997) even if maybe this specificity would be rather face-modality dependent (Robotham
et al., 2011), or anger detection for instance is more impaired in Parkinson patients (Lawrence et al.,
2007). Autistic patients show social deficit in emotion perception and expression in possible relation
with an alteration of the imitation functions (Williams et al., 2001). Notably the specificity of deficits
observed in different patients with different pathologies shows the relative specificity of neural
substrates toward the diverse emotions or their modalities. The study of disrupted emotional
8
mechanisms takes part in the understanding of various pathologies. Conversely acute, strong or
repeated emotional events could also induce atypical reactions and long-lasting pathological states.
Indeed the alterations found in post traumatic stress disorder patients are related to previous
experience or witness of a highly traumatic event conveying highly negative emotional valence.
4. Emotion modalities
In addition to the different aspects or sub-functions and the different classes of emotions, a
partition in modality is also possible in order to appreciate the complexity of emotions. We will focus
on visual perception of emotions and more particularly in emotion in faces and in words.
4.1. The different emotional modalities
Emotions in a wide sense of communication are conveyed via diverse vectors. In fact the
wide concept of emotions is often combined with a particular modality. The term emotion could
hardly be separated from a modality, from the substance containing emotion. Thus one speaks about
emotional face, emotional voice, emotional gesture, emotional word and semantic, emotional music,
emotional scene or emotional situation. These are the main modalities possibly conveying emotional
content. As in the case of the functional aspects of emotion – i.e. perception, interpretation, and
expression – the different modalities share common points in their general neuropsychological
functioning and they also differ in their incidences towards the emotion domain as they respond to
specific schemes of execution. For instance gesture and facial expression would show close mimicry
models and it is not clear if mirroring mechanisms are similar or even present in vocal modality.
Mimicry is the automatic imitation of others’ gestures. Mirror processing refers to the functioning of
the mirror neurons that show activity during production and perception of similar specific actions..
Differently, codes for emotion in music are absolutely different from those in emotional scenes but
could have similarities with vocal codes of emotion expression for instance – e.g. drastic changes in
tonality and amplitude relate generally to a high emotional arousal. Generic aspects in emotional
modalities would be also transposable to the strict level of the neural substrate. Some cerebral
structures have a general involvement and show inter-modality response as for instance the
amygdala or the insular cortex that could be implied in response to emotional content in facial or
musical modality (Hseih et al., 2012). The idea of a supramodal functioning in emotion processing is
corroborate by numerous studies (e.g. Peelen et al., 2010). Furthermore, these cognitive similarities
are also markers of the affective state and of the general responsiveness to emotions (Blanchette
and Richards, 2013; Wilson-Mendenhall et al., 2013). Conversely the emotional overlay may interact
9
with the considered modality and therefore shares some specificity of the modality. Hence other
cerebral structures are modality specific but answer also as a function of differences in emotionality.
4.2. Focus on two modalities, faces and words
Faces and written words are the two modalities studied in the present work. They are also
the most studied in neuroscience. As an index, the request in PubMed search engine of the National
Center for Biotechnology Information with the term emotional associated with face reaches 6361
results, and with word it reaches 1970. It reaches 1256 with emotional voice or 327 with emotional
scene. Faces and words constitute also major means of emotional communications in humans.
The face is an important vector of emotional interaction. The intuitive association between emotion
and face is certainly due to the automatic, rapid and implicit functioning of facial expression
(Cacioppo et al., 1986). Thinking about any class or category of emotions reminds a smile or a frown,
a pout or a tear. Conversely facial expression is an automatic way to express a current emotional
affect. The ideological link between the concepts of emotion and of facial expression might be due to
the role of facial mimicry in facial emotion recognition (Ponari et al., 2012). Numerous studies have
investigated the role of facial mimicry in the emotion thematic but its requirement for a proper
recognition is sometimes criticised. Absence of facial movement has been shown to be effectless on
emotion recognition in a study with facial paralysed patients (Rives Bogart and Matsumoto, 2010).
This last study might be yet biased by possible plasticity processes engaged in these patient
populations to achieve emotion recognition, moreover the absence of movement in these patients
does not necessarily imply a deficit of activation of corresponding imprint in the cerebral motor
system. Humans are expert in face detection and recognition, and they are extremely good at
detection of fine emotions (Marneweck et al., 2013). The mechanisms of facial expression behaviour
in humans seem to take anchor in ancient phylogeny as similar processes have been observed in
monkeys (Lefrou, 1956; Fernandez-Carriba et al., 2002; Parr and Heintz, 2009).
Articulated language is the main way to communicate explicit concepts and complex thoughts with
others. Hence verbal material has also become a major vector for expressing emotions alongside
with the other modalities. Oral and written words differ in physical feature, sensorial entries and for
some parts of the cerebral treatment of course; but they share also common codes with semantic
and lexical levels of processing. The two expression systems in language, oral and written, rely on
very different sensorial modalities and are constrained with the properties of these modalities. These
differences could be due to very low level properties. For instance, a written word is read letter by
letter or grapheme by grapheme but possibly is instantly seen and recognised as an entire word. On
10
the contrary the recognition of spoken words can’t be faster than the flux of the speaker. However
anticipatory mechanisms are set in play in order to help recognition of end of written words, and
words in a sentence while reading, and it is also true in the oral modality within sequence of
phonemes and words; without necessarily being conveyed by the exact same mechanisms.
Concerning emotional aspects, the emotional tint is generally instantly detected in spoken language
by the use of particular tone and rhythm in the voice while written words transmit emotions with
their literal meaning – plus the punctuation. These are some of the intrinsic differences between oral
and written modalities. Hence written language, out of any form of the graphemes, punctuation and
sentence construction, express its main emotional connotation by the meaning of the words. The
signified can be linked to a lexical field of particular emotion and affect; the signifier is hence
considered as an emotional word. Emotional words represent a specific category among other words
in term of cognitive and cerebral processing. They are recognised more easily and more accurately
(Graves et al., 1981; Strauss, 1983; Vinson et al., 2013; Scott Et al., 2013). Their interference with
reading can be studied in emotional stroop tasks (Williams et al., 1996) and is though debated
(Dresler et al., 2009). They engage also additional neural structures compared to neutral words
(Isenberg et al., 1999; Hamann and Mao, 2002; Ortigue et al., 2004). This point was investigated in
the second study presented in this report (Rochas et al., 2014).
5. Neural structures in emotion and potential targets
5.1. Emotion structures and organisation
The emotional neural substrates, yet not completely known and understood, constitute a
large bunch of potential targets for TMS. The hypothalamus was one of the first cerebral structures
suggested to be implicated in emotions (Bard, 1928; Bard and Rioch, 1937). In the 1930’s, the
American neuroanatomist James Papez proposed a cerebral model of the entry of emotional stimuli.
This system, termed the “Papez circuit”, includes the thalamus, the hypothalamus, the fornix and the
cingulate cortex (Papez, 1937). It introduced a vision of upstream and downstream regulations
between the different cerebral structures. Adapted from Papez circuit, neuroscientist and physician
Paul D. MacLean described the limbic system (Maclean, 1952) as a visceral brain and was supposed
to integrate visceral and somatic signals in response to external emotional stimulation. Maclean’s
model supported the idea of the experience of emotion via the body. Throughout the subsequent
experiments, rather minor role in emotions was attributed to the diverse structures of the limbic
system in favour of other functions – e.g. the hippocampus and declarative memory. The localisation
of emotional processing has been diffused to the entire brain in many different studies. However
11
some of the described structures, extensively investigated in affective neuroscience, have remained
important actors of emotions. The amygdalae, the striatum, the hypothalamus, or the cingulate
cortex are indeed some of the major actors of the emotional brain. Additionally and in interaction
with the limbic system, cortical and subcortical regions can also show specific implication in the
processing of the emotions according to the different dimensions (Adolphs et al., 2003; Posamentier
and Abdi, 2003; Posner et al. 2008).
First based on neuropsychology and then on neuroimaging, affective neuroscience lead also to the
emergence of various hypothesises of brain lateralisation. The valence hypothesis has posed a
dichotomy of the brain between positive emotions being processed in the left hemisphere and
negative emotions in the right hemisphere. The idea has been raised in the late 1950s and early
1960s after observations of opposite affective reaction following injection of sodium Amytal –
provoking the inactivation of the brain tissue – in unilateral carotid vascularising the left or the right
hemisphere (Terzian and Cecotto, 1959; Alemà and Donini, 1960; Perria et al., 1961). Different
behavioural studies also supported the valence hypothesis (Reuter-Lorenz and Davidson, 1981). This
lateralised distinction would be more robust in the prefrontal cortex (Tucker et al., 1981; Davidson et
Fox, 1982).
In parallel a right hemisphere hypothesis proposed an overall dominance of the right hemisphere in
the general emotional treatment. This assumption has been first deducted from lesion patients
(Gianotti, 1969; Gianotti, 1972; Borod et al., 1988) who showed drastic different reactions according
to the side of brain damage. Left lesion patients expressed depressed and catastrophic reactions
while right lesion patients showed indifferent behaviour evoking rather a loss of emotional sense
consequently to the alteration of the right side. Then the right hemisphere dominance in emotion
has also been showed in numerous modalities. The advantage has been found in verbal modality
(Borod et al., 1998), in auditory modality (Erhan et al., 1998), in facial expression (Sackeim et al.,
1978; Mandal et al., 1996).
Psychologist and psychiatrist Richard Davidson later proposed a revised version of the valence
hypothesis and stated that it was applicable for emotional response only then, he turned it in an
approach-withdrawal model (Davidson, 1998; Workman et al., 2000; Davidson, 2003). The
motivational hypothesis was, inter alia, proposed to explain some incongruities observed with the
valence model of lateralisation for anger being negative as sadness or fear but being approach
inductor as happiness.
Across different affective neuroscience studies, the inconsistencies of some results lead from the
right hypothesis model to the valence hypothesis model then in favour of the motivational model.
12
Some authors also suggested an integration of the different models of lateralisation with proposition
that the right prefrontal cortex processes preferentially negative valence emotions, while the right
parietal lobe is involved according to the arousal, and the left prefrontal structures are biased toward
positive valences (Heller et al., 1997). As the hypotheses of lateralisation could not explain every
situations, one proposed that the lateralisation according to the valence or the approach were
dependent on the considered cerebral structure (Wager et al., 2003) or that arousal and valence
dimensions would rely on distinct neural networks (Posner et al. 2008). Consequently there would
not be emotional lateralisation per se and it appears more complex than it was supposed to be in the
former theories. The whole emotional system is distributed through multiple cortical and subcortical
structures but showing distinct specificity according to the different emotions and modalities, plus
each system could present a different lateralisation. Moreover the lateralisation is also variable
depending on individuals and for instance according to the factor gender (Wager et al., 2003). These
findings lead to conceptualise differently the cerebral model of emotions.
Distributed model of emotions were developed without a priori on structures but rather with a focus
on a specific condition, modality or emotion class. The key actors of emotional brain are first
represented by the amygdalae which are implicated in numerous of emotional aspects as fear
conditioning (LeDoux, 1993; Bechara et al., 1995), facial expression recognition (Morris et al., 1996;
Morris et al., 1998; Adoplhs et al., 2005), unconscious processing (Gianotti, 2012) or emotional
memory (Cahill et al., 1996; Hamann et al., 1999). Connected to the amygdalae, the hypothalamus
takes also part as a main actor of emotion cognition. As amygdalae, the hypothalamus shows
activation by viewing of positive or negative emotional pictures (Karlsson et al., 2010). Its electrical
stimulation in animal models has shown bodily reactions of recognisable angry state (Vergnes and
Karli, 1970). As amygdalae and hypothalamus do, the prefrontal cortex region is also involved in
reward cognition, though it is implicated in pure emotion response as well. Prefrontal cortex has first
been suspected to be an emotional control region with the case of the famous patient Phineas Gage
who had drastically changed his behaviour after an iron bar came across his prefrontal cortex in an
accidental explosion in 1848 – yet he had surprisingly not passed away. Particularly he showed
disinhibited behaviour and angry mood (Harlow, 1868). Later studies in early neuroscience looked at
the role of this region in emotional processing. Following a current of thought of the integration of
bodily signals in emotion processing, Antonio Damasio developed the somatic marker hypothesis
that states a monitoring of signals from the body and physiological reactions in the interpretation of
the emotional events (Damasio, 1996). More precisely, this interoceptive treatment would be
executed in the ventromedial prefrontal cortex. Notably the prefrontal cortex represents a key
structure in the interaction of emotion and decision making. Interestingly some areas seem to act
13
like associative emotional structures as it is the case for the orbitofrontal cortex which collects
information from the different sensory modalities (Rolls, 2004). Neuroimaging also showed
interactions in activity between different modalities. Cerebral activity responding to fearful
emotional language material was for instance modulated by simultaneous presentation of neutral
face material in an fMRI study (Willems et al., 2011). We will focus our investigations on particular
cognitive aspects of these two last modalities.
5.2. Neural treatment of facial emotion expressions
As any other emotional features, emotional faces elicit particular brain states in their
processing compared with other modalities (Krolak-Salmon et al., 2001). However emotional face
recognition is of particular daily importance for a normal social life. Facial expression recognition can
be detected really quickly as shown by specific components found in electrophysiology studies (Eimer
and Holmes, 2002; Batty and Taylor, 2003; Pourtois et al., 2004; Krolak-Salmon et al., 2006; Rellecke
et al., 2011; Calvo et al., 2013). When investigating facial expression processing, the studies focus on
difference of cerebral structures according to the different emotion classes often based on the basic
emotions defined by Paul Ekman (see figure 1). This function is ensured by a combination of parts in
the brain devoted to emotion or faces. In fact comparative review showed a lot of similarity between
the facial expression and the general emotional processing (Vytal and Hamann, 2010). Although
among the specialised structures of facial recognition, the fusiform gyrus, containing the fusiform
face area, (Dolan et al., 1996; Vytal and Hamann, 2010) and the superior temporal sulcus (Fried et al.,
1982; Narumoto et al., 2001) are also reported to have a role in facial expression processing. The
implication of the amygdalae already known for aversive conditioning has been shown in the
recognition of facial expression of fear (Morris et al., 1996; Phillips et al., 1998; Krolak- Salmon et al.,
2004). A recent meta-analysis of PET and fMRI studies about the regions activated during the
processing of emotional stimuli confirmed the implication of both amygdalae in fear face processing
and showed additionally consistent activations of the right cerebellum, the right insula (BA 13), or
the left fusiform gyrus (Vytal and Hamann, 2010). More generally the amygdala is not restricted to
fear. Sadness when detected on a face is also partly processed in the left amygdala and in the right
inferior and middle temporal gyri as shown in a PET study (Blair et al., 1999). In this same study, the
anterior cingulate cortex showed activation for both increasing sadness and anger while anger
expression showed additionally specific orbitofrontal cortex and striatum activations (for review
Posamentier and Abdi, 2003). Facial expression of disgust is closely related with the specific
activation of the anterior insula (Phillips et al., 1997; Krolak-Salmon et al., 2003) and additionally the
caudate-putamen (Phillips et al., 1998). The expression of surprise has been studied in less extent.
14
This facial expression is close to the expression of fear. Although its detection is related to the
activations of amygdalae as for fear detection, activations are also found in the medial prefrontal
cortex (Kim et al., 2003). These results would be dependent on subjective interpretation and surprise
could be considered to be a transitory state towards another emotion. The results found for surprise
would be related to this context rather than expressing an independent and particular state (Kim et
al., 2003; 2004). In fact the activation responses to surprised face are correlated with the subjective
judgment of valence (Kim et al., 2003). Note that similar response to negative assessments was
found in the amygdala for neutral faces as well (Blasi et al., 2009) linking the amygdala response to a
motivational aspect rather than independent effect related to specific types of emotion. The metaanalysis of Vytal and Hamann reported activations of the right superior temporal gyrus (BA 22) and
left anterior cingulate cortex (BA 24) in the processing of happiness in facial expression (Vytal and
Hamann, 2010). The early activation of pre-SMA was clearly shown in a combined fMRI-EEG study for
facial emotion recognition at least for happiness and sadness (Seitz et al., 2008). The same metaanalysis study of Vytal and Hamann, acknowledged activations in the left medial frontal gyrus, the
right inferior frontal gyrus and inter alia the left caudate head for sad face and in the left inferior
frontal gyrus, in the right parahippocampal gyrus or in the left fusiform gyrus for the processing of
anger. Interestingly Sex differences exist at the neural functioning level in response to emotional
stimuli. Experimentation with emotional facial expressions showed correlation in insula activity in
women possibly reflecting interoception and it showed correlation with the visual cortex in men
possibly reflecting a more visual representation (Moriguchi et al., 2013). Further than the attribution
of a bunch of structures to a given emotion, an experiment on non-human face materials tells us that
there is no potentiation to read emotion on human face but rather the detection of facial expression
on non-human face would be an extension (Chammat et al., 2010).
The redundancy of certain structures in processing of different emotions reveals an integrative
functioning of these mechanisms. Hence the detection processes for the different facial expressions
rely on a common system but in the same time differ also from each other with some specific
activations. The activity within the diverse implicated structures would be ultimately integrated to
form a message from the whole bunch of processed information and to understand the emotion
currently perceived. In the same time it is not always clear if the activations reported in
neuroimaging are necessary for the simple detection of expressions. Moreover the majority of
studies focus their results on positive activations, but following an integrated view, the non
activation of some cerebral structures might have significant meaning and role in the definition of
the perceived facial expression. For instance, loss of amygdala lead a patient to misinterpret fearful
and angry faces in rather happy faces (Sato et al., 2002); this patient showed a general bias toward
15
happiness when asked to identify facial expressions. The integration of the significant neural
messages occurs also with the integration of the activities in other structures related to other aspects
of facial detection – i.e. face shape, direction of gaze, gender, and other subjective judgments as
attractivity or trustiness – in order to form a complete representation of faces.
Out of an emotional categorisation the large bunch of structures implicated in the treatment of facial
emotion recognition can be divided in two categories, according to their functional purpose (van der
Gaag et al., 2007). In fact, the “traditional” and well-known emotional structures as the amygdala
and the insula could be responsible for the empathic reaction to emotional face perception. The
empathy is the simulation of the feelings and the sensory compound related to another’s emotion by
an individual. In contrast, the motor and premotor areas could process the mimicry reaction to
emotional face perception. These regions could simulate motor expression of another’s emotion.
These two mechanisms might have complementary roles in the emotion processing. Interestingly
motor and somatosensory areas showed common responses to both expression and perception of
pleasant facial emotional stimuli (Hennenlotter et al., 2005). In studies on epileptic patients, the DBS
of the pre-SMA provoked uncontrolled reaction of happiness and joy feeling (Fried et al., 1998;
Krolak-Salmon et al., 2006). Additionally the pre-SMA showed also a specific evoked potential to
happy face presentation in an implanted patient (Krolak-Salmon et al., 2006). An fMRI study has
shown activation of this area in response to both voluntary imitation and facial emotion recognition
(Carr et al., 2003). In a review paper, Kober and colleagues discussed the role of the premotor areas
rather in a direction of action preparation in response to emotions and not directly in emotion
recognition (Kober et al., 2008). Our TMS experiment on the left pre-SMA will potentially answer to
this divergence.
Some authors raised also criticism about general results of studies on emotions in regard of the
modulatory aspects of the treatment of emotions (Posner et al., 2008). The low reproducibility of
some of the results and the limited number of regions implicated in facial emotion recognition reflect
possibly some technical limitations in the paradigms used in the study of such a complex function. In
imaging, the frequent utilisation of neutral stimuli as contrast to reveal processing of the emotional
ones is questionable. The neutral stimuli are not necessary devoid from any emotional aspects and
could be apt to be interpreted as ambiguous emotional expression. Anyway neutrality could engage
emotional processing and this may create noise and interferences in the studied contrasts.
5.3. Neural treatment of emotional language material
16
Emotions constitute special material in general language. In reading, written words denoting
emotional valence or arousal lead to modified performances when compared to neutral words
(Graves et al., 1981; Schacht and Sommer 2009; Kissler and Koessler 2011; Skrandies, 2013; Landis,
2006; Rochas et al., 2014) and more interestingly their processing elicits modification in brain activity
as shown in neuroimaging (Ortigue et al. 2004; Frühholz et al. 2011; Ponz et al. 2013; Herbert et al.,
2011; Palazova et al., 2013; Rochas et al., 2014). Compared to the neutral words, increased
activations or divergent networks are engaged. The amygdalae were found to be implicated in many
studies on emotional words (Isenberg et al., 1999; Hamann et Tao, 2002; Straube et al., 2011; Kanske
and Kotz, 2011, Citron, 2012). In the same manner, the insula or the prefrontal network were also
found to be part of emotional words processing (Lewis et al., 2007; Herbert et al., 2011; Balconi and
Ferrari, 2012; Ponz et al. 2013). Considering the panel of structures involved in the processing of
emotional words, it is noteworthy that the classic emotion-related cerebral network is also
implicated by word reading. Also the activations found during emotional word reading can be
modified by independent high-order factors as self-reference for instance (Herbert et al., 2011); the
activations would be more related to emotional mechanisms rather than word processing. This
implication might reflect the consecutive activation of the emotional network after the semantic
analysis of the stimulus, however an independent computation of emotionality directly from word
material is also considered. The early involvement of some areas in the specific processing of
emotionality of words argues in favour of an independent processing in parallel with the pure
semantic processing. Interestingly, thanks to high temporal resolution of the techniques used by
Ponz and colleagues, their results are discussed in favour of an early and dedicated emotional
analysis of words in the insula additionally to the semantic analysis. In fact many studies in
electrophysiology showed particular early divergence between neutral and emotional word material
(e.g. Palazova et al., 2013) and frequently described as a specific early posterior negativity
component (Citron, 2012). The specificity of emotionality processing in words can thus be identified
as early as the visual cortex (Frühholz et al. 2011). Note that this early phenomenon in emotional
word processing can be dependent on the considered word class (Citron, 2012). More interestingly it
is sometimes compared to the automatic attentional focus described in other modalities for
emotional stimuli (Sommer and Schacht, 2009; Palazova et al., 2013). These arguments are
inconsistent with the traditional view of a cascade of decoding from visual feature to semantic
meaning. Authors argue rather in the direction of the potentiation of the reading network by
emotional words (Keuper et al., 2014). Semantic features can also influence the emotional processing
(Kanske and Kotz, 2011).
17
The studies investigating the neural bases of emotional word processing rarely dismantle the
emotions between the basic classes. Although they study often differentiate the emotions along
dimensions of arousal or valence – i.e. neutral vs. emotional, positive vs. negative. In fact Lewis and
colleagues (Lewis et al., 2007) decomposed the activations along the two dimensions with the
orbitofrontal and the ventral anterior cingulate cortices according to valence of the words, and the
amygdala, the anterior insula and the pallidum responding along the arousal dimension. Another
fMRI study showed responses in the left extra-striate cortex for emotional words compared to
neutral words and an increased response in the right insula and superior temporal gyrus for positive
high-arousal and negative low-arousal words and similar results for positive high versus low-arousal
(Citron et al., 2014). Authors discussed the dimension-dependent activation of the insula in term of
integration of the approach-withdrawal tendencies, however again a large network is implied in
response to emotional words presentation.
5.4. Right hemisphere in reading of emotional words
The right hemisphere hypothesis claiming for a right hemisphere dominance in emotions has
been challenged with emergence of distributed models. Emotional words increase the activity in the
classical language-related regions in the left hemisphere. However the right hemisphere has clearly
showed evidences for its role in diverse aspects of emotions. In language and word reading, data
from aphasic and split brain patients first showed implication of the right hemisphere (reviewed in
Landis, 2006). Thus advantages for emotional words were observed in patients with lesion of the left
hemisphere (Landis et al. 1982; Landis et al. 1983; Reuterskiöld 1991). Conversely, studies in patients
with fronto-temporal damages in the right hemisphere reported difficulties with emotional word
(Borod et al. 1992; Lalande et al. 1992; Cicero et al. 1999) or social disturbance (Perry et al., 2001; Liu
et al., 2004). Processing of emotional words is found in the right hemisphere also in healthy
participants. First behavioural study showed an advantage for the right hemisphere to detect
emotional words (Graves et al., 1981; Ortigue et al., 2004; Smith and Bulman-Fleming, 2005) at least
under certain conditions (Mohr et al., 2005). Neuroimaging studies have also implicated the right
hemisphere in emotional processing (Frühholz et al. 2011; Ponz et al. 2013). According to the review
of Abassi and colleagues, the two hemispheres act in cooperation in the processing of words and
emotional words (Abassi et al., 2011) and it is what our second TMS study proves. However the
authors argued in their review also in favour of a secondary involvement of the right hemisphere
when the left presents an automatic mechanism of word detection. This latter opinion is not in
concordance with our EEG experiment that states for bilateral involvement of the temporoparietal
junctions with specific activations in the right hemisphere (Rochas et al., 2014). The temporal lobe,
18
seat of the Wernicke area, is an important region for word comprehension. Its right homologous can
also be implied in various aspects of language processing (Stoeckel et al. 2009; Binder et al. 1999;
Bonner et al. 2013; Buchsbaum et al. 2001; Vigneau et al. 2011; Diaz and Hogstrom 2011; Kuchinke et
al. 2005; Graves et al. 2010). In our study we used a simultaneous bilateral lexical decision in order to
dismantle the intervention of the two hemispheres and to “force” the right hemisphere to directly
process pure word material. The specificity of our task has stressed the need of a localising phase
with a prior EEG experiment. Afterwards it has been chosen a target in accordance with our
expectations. It was in the right hemisphere and susceptible to be emotion-related, at the junction of
the temporal lobe for word comprehension and the parietal lobe in multimodal associative areas.
III. TRANSCRANIAL MAGNETIC STIMULATION AND APPLICATIONS
Although we will be interested in the short-term and direct effects of TMS in this thesis work,
the general functioning and the different aspects of the technique will be reviewed. The use of TMS
in fundamental neuroscience constitutes our main focus. However interesting applications in clinics
will be also reviewed. As a brain stimulation technique, safety guidance and potential risks of TMS
will also be discussed.
1. History of TMS: from bioelectricity to modern TMS
The early development of transcranial magnetic stimulation has been carried by progresses
in electricity and physiology of the motor system rather than by brain research. The transcranial
magnetic stimulation principle has taken its roots 250 years ago when the Italian physicist and
physician Luigi Galvani in 1771 managed to stimulate animal muscles with a spark. Galvani concluded
from this experiment that a bioelectricity or “animal electricity” must exist (Galvani, 1791). In the
early 1840’s, another Italian physician, Carlo Mateucci, recorded electrical activity from muscles and
its transmission through the nerves (Mateucci, 1844). He proved that nerves were sensitive to
electricity. At a contemporary period but in another domain, continuing the works from Hans
Christian Ørsted, the English scientist Michael Faraday established the first principles of
electromagnetic induction in 1831 (Faraday, 1839-1855). This part of his work has been formulated
as the Faraday’s Law by James Clerk Maxwell (James Clerk Maxwell, 1861). In 1870, the German
neuroscientist Eduard Hitzig, assisted by the anatomist Gustav Fritsch, induced movement in
different muscles of a dog by applying electrical current with an electrode on different locations of an
area of the cortical brain that they called the “motor band” (Hitzig & Fritsch, 1870). In the 1870’s, the
Scottish neurologist David Ferrier meanwhile stimulated with electricity and injured the cortex of
many species in order to map brain functions. Noteworthy he used faradic stimulations in his
19
experiments (Ferrier, 1876). In 1874, the American physician Robert Bartholow applied such faradic
electrical currents on the dura of a human trepanated patient, who unfortunately died following the
experiment, reproducing some of the previous animal results in a human being (Bartholow, 1874). It
is in 1896 that the French physicist and physician Arsène d’Arsonval combined the advances of
electromagnetism and electrical stimulation. Interestingly, he demonstrated the appearance of some
visual sensations that he called “magnetophosphenes” when the head of an individual was placed in
a time-varying electromagnetic field (see figure 2 A). In fact this perceptual effect was later explained
by the excitation of the retinal cells which are very sensitive to rapid changes of the magnetic field,
rather than the cortical cells. Nevertheless, it constituted the first experimental magnetic stimulation
of nerve cells. These induced visual sensations have been replicated during the XX th century with
varying types of solenoids until the emergence of more powerful current generators and the
improvement of magnetic coils. In 1965, R. Bickford and B. Fremming developed a stimulator capable
of generating high frequency sinusoidal magnetic pulses and achieved the stimulation of human
nerves (Bickford & Fremming, 1965). Following the work of Merton and Morton who used noninvasive scalp electrodes to stimulate the brain electrically thanks to rapid currents, Anthony Barker
and colleagues succeeded in the first magnetic stimulation of the cerebral cortex with a coil
conveying rapid current placed at the head surface in the 80’s (Barker and Freeston, 1985; Barker,
Freeston, Jalinous, Merton, and Morton, 1985; Barker, Jalinous, and Freeston, 1985; see figure 2 B).
Figure 2: A. from Brodier, 1897: picture of a device of electric autoconduction (or Darsonvalisation) adapted from the
work of d’Arsonval on electromagnetism in humans.
B. Reza Jalinous, Ian Freeston and Antony Barker with the stimulator of 1985 first succeeding in stimulation of the
cerebral cortex.
20
2. Principles of TMS
2.1 Mechanisms of stimulation: from electric current to cell depolarisation
In modern magnetic stimulation, the electric field responsible for the stimulation of the
nervous tissue is induced by a strong and rapidly changing magnetic field according to the MaxwellFaraday equation. The induced magnetic field is up to several Tesla strong and is generated by brief
currents of several kiloAmperes (kA). The magnetic field is able to penetrate through the skull
without attenuation – contrary to the electric field, which is attenuated by the high impedances of
bone and skin. To achieve relevant stimulations of the cerebral cortex with direct electric stimulation,
the electric field would have to be rather high (low current with high voltage, e.g. 20mA and 250V)
and would not be tolerated by the subject. The magnetic field only attenuates with distance. Thus it
is particularly interesting for the remote stimulation of the cerebral cortex from outside of the head.
In the case of TMS, this particular magnetic field is generated by a very rapid and brief high voltage
current injected into a wire coil. The coil acts as an emitter and the cerebral tissue acts as a receptor
during the stimulation pulse. The electric current in the coil induces a time-varying magnetic field
perpendicularly to the direction of the current according to the Ampère's circuital law (with
Maxwell's correction) – i.e. perpendicularly to the coil wire plan. The slope of the current change is
important in the stimulation efficiency – i.e. it has to be sharp to generate a strong magnetic field.
The changing magnetic field induces in return a changing electric field according to Faraday’s law of
induction and then an electromotive force in an adjacent electric conductor translated as micro
current flows in the brain tissue. These induced micro current flows are directed perpendicularly to
the direction of the magnetic field and therefore parallel but opposite to the electric current
direction of the coil wires. The TMS does not add electric energy in the head, but it creates remote
electric current flows. Note that this electromotive force emf is proportional to the variation of
current that goes through the coil in time and to the proper inductance of the coil (defined according
to the equation: emf = - L x d(I)/d(t)). The short indirect micro current flows change the polarity
around the cell membrane in a transient manner. Thus, the membrane resting potential is changed.
The changes in ambient polarity, depolarisation or hyperpolarisation, around the neural cell induce
movements of ions via the passive ion channels on the cell membrane. The change of polarity of the
nerve cell eventually leads to activate the voltage-dependent ion channels of its membrane and
generate the depolarising action potential of the cell or hyperpolarisation. This is the operative
principle of the “activating function” of the TMS (Barker, 2002). The effect of the driven currents is
more likely to stimulate cells on their axons, where they terminate or where they sharply bend
(Amassian et al., 1992; Maccabee et al., 1993; Nagarajan et al., 1993) due to the interaction of the
21
ambient electric current with these special configurations of the membranes. The pyramidal cells are
activated directly on their axon stem representing the direct wave of activation, or through synaptic
communication representing the indirect wave (Kenrell & Chien-Ping, 1967; Day et al., 1989). Thus
the supra-threshold magnetic stimulation might result from action potentials elicited by the
interaction between the induced brief current flows and the cell membranes of the stimulated cells.
This mechanism is responsible for the online effect of TMS. The offline effects of TMS might be
caused by long-term changes in the excitability of the stimulated cell, due to supra or sub-threshold
repeated stimulations, as well as through remote effects by modulations of connected cells through
networks (Thickbroom et al., 2007).
2.2 Under the coil: distribution of currents
According to Faraday’s law, the magnetic field is created perpendicularly to the plane of the
coil loop and without any attenuation by the layers separating the coil from the cerebral tissue. The
maximum of current density at the cortical level is obviously related to the magnetic field distribution.
The maximum probability of triggering depolarisations of neuronal axons should be encountered in
the zone underlying the coil where the current density is the highest. This assumption has been
verified in several studies with different methods partly reviewed by Wagner and colleagues (Wagner
et al., 2009; see figure 3). Indeed, phantom studies (Tay et al., 1989; Yunokuchi et al., 1998), depth
electrode recordings in humans and animals (Tay et al., 1989; Lisanby, 1998; Wagner et al., 2004),
imaging studies (Bohning et al., 1997; Valero-Cabré et al., 2005, 2007; or indirectly Wassermann et al.,
1996), and electromagnetic modelling (Wagner et al., 2009) have been used to investigate and
evaluate the impact of TMS on the cortex. The various results showed a maximum of current density
and effect of isolated pulses in the target-area being in the induced magnetic field maxima (Tay et al.,
1989; Bohning et al., 1997; Wagner et al., 2004). This maximum current density localised relative to
the coil is used to target the cortical area to be stimulated. However, due to possible inhomogeneity
of the tissue and of the cerebral cortex, the distribution of current density could be deviated from
what is expected in a homogeneous model – i.e. a spherical or plane superposition of homogeneous
layers. Thus the maximum of current density may not be distributed in a homogeneous manner as it
follows the conductivity of the different tissues and their arrangement – i.e. gyri and sulci. As shown
by Opitz and colleagues, the distribution of the induced electric field is higher when the current flows
are perpendicular to the considered gyrus (Opitz et al., 2011). In addition they found that the gray
and white matters do not concentrate the electric field in the same manner. The electric field is
stronger at the crowns and lips in the gray matter while it could penetrate notably deeper in the
white matter. Note that the induced electric field is also critically contained by the high resistivity of
22
Figure 3 from Wagner et al., 2009: Multiple methods exist to evaluate the stimulating fields, including A. Phantom
model recordings have been made with probes that are placed within containers of various geometries containing
saline or other materials used to model the human head. B. Depth electrode recordings have been made of the current
densities or induced voltage gradients in human and animal subjects. C. Imagining studies have been used to provide
field information (adapted from Nagarajan S, Durand DM, and Warman EN. 1993. Effects of induced electric fields on
finite neuronal structures: a simulation study. IEEE Transactions on Biomedical Engineering, 40: 1175–1188.). D.
Electromagnetic models are the primary method used to predict the stimulating field distributions.
the skin and bone. Moreover the observation of maximal effect underlying the stimulator coil is
nuanced in the case of repeated stimulation and long-term effects with some scattered effect
manifestations in the brain. Interestingly these projected effects are observed in the connected area
receiving efferent projections from the stimulated cortex area (Valero-Cabré et al., 2005). While the
magnetic field is not disrupted by the skin and bone layers, it decreases rapidly with distance. Thus to
achieve the stimulation of moderately deep areas of nervous tissue (like sulci) and deep structures,
the output power has generally to be strongly increased with a detrimental effect on focality of the
stimulation, on coil cooling, and on the comfort of the subject. At least with the classical systems and
coils commonly used, there is a trade-off between focality and depth. TMS will therefore achieve a
depth of stimulation that is limited to the cerebral cortex (excluding the likely projected effects).
23
2.3 Different current waveforms and their implications
Capable of generating a varying strong magnetic field, the TMS systems are composed as an
RLC circuit. A capacitor and a stimulating coil together with their own resistances and the resistance
of the circuit constitute the elements of the RLC-like circuit. In addition, a thyristor switch commands
the triggering of the pulses. This construction permits to generate rapid currents – i.e. high variation
of current in a reduced time. The stimulation is as more efficient as the rise time of the magnetic
field is quick for the same stored energy (Barker et al., 1991). Modern TMS systems are built to
release the current in their coils as fast as possible, leading to a reduced storage of energy in the
electric system for comparable stimulations. As a consequence of using a weaker energy, the power
supply, the current peak, and the coil heating can also be reduced. However the rapid current
launched in the coil produces a strong sound due to the fast magnetic-induced movements of the
wires making them intershock in the coil sheath (Tringali et al., 2013). The TMS devices can produce
different types of currents. The most common types are either monophasic or biphasic currents. A
monophasic current is characterised by a unidirectional flow of electric current in the wires of the
coil in one phase, whereas a biphasic current is characterised by two consecutive phases with
opposed currents. In addition of moderate resistances, the generator system for biphasic pulses also
includes a diode that in a second phase allows for the return of a part of the current in the opposite
direction to the initial current from the capacitor. In the monophasic pulse generator, the diode and
resistor shunt the current leading to annihilate the second phase (Barker et al., 1987). A current
induced by a monophasic pulse (less than 100 µs) consists in a positive phase representing the first
quarter of the revolution of the sinusoid of the current induced by a biphasic pulse (300 µs) (Sommer
et al., 2006). The physiological effects of monophasic or biphasic currents are slightly different (Di
Lazzaro et al. 2001). The biphasic TMS is more effective than monophasic TMS in the activation of the
nervous tissue for the same stored energy of stimulation or similar upstroke of the pulses (Claus et al.,
1990; Brasil-Neto et al., 1992; Kammer et al., 2001, Arai et al., 2005). Monophasic and biphasic
currents present also opposite current directions of the first phase for the most effective stimulation
– i.e. the orientations of the coil would be opposite (Maccabeh et al., 1998; Kammer et al., 2001).
Authors explained this inversion by the fact that, in biphasic pulses, the first hyperpolarising phase
might potentiate the Na+ channels for the following depolarising phase, or that the second phase
length (half of the revolution, twice longer than the upstroke phase) is more propitious to depolarise
neural cells (others than for the upstroke phase). In addition, the latency of an MEP is also shorter for
biphasic compared to monophasic stimulations (Sommer et al. 2006). According to their biologic
effects and their respective operation, they are designed for different preferential utilisations.
24
Monophasic rTMS seems to be more efficient than biphasic for facilitation effects (Tings et al., 2005;
Arai et al., 2005) or long-lasting effects (Sommer et al., 2002). The superiority of monophasic
stimulation in rTMS protocols might be partly due to the better potency to change the excitability
during the stimulation (Arai et al. 2005). However, the differences of efficiency for the induction of
long lasting effects are difficult to interpret as the stored current intensities were higher for
monophasic system – i.e. adapted to the motor threshold. The monophasic system also consumes
more energy and induces more coil heating, which are important issues for high frequency protocols.
In our studies the repletion of pulses was not problematic as the train of pulses were separated by
relatively long pauses. However pulses delivered with the machines were biphasic and provided
maximum efficiency for event-related TMS.
The decisive factor in the stimulation of excitable cells is not the amplitude of induced current but
rather the charge that is generated by this current at the cell membrane level. The membrane cell is
being considered as an RLC system. As suggested by Corthout and colleagues (Corthout et al., 2001),
the opposite effect of two types of stimulator was attributed to the time shift in the appearance of
maximal charge at the cell membrane level. Charge changing is a matter of current change, and
consequently the electromotive force, according to the time. Thereby, the emphasis has to be on the
dynamic of the charge change.
The length of the induced current is also of critical importance in the sense that the current change
around a nervous cell has to last a sufficient time to allow the ions to traverse the cell membrane.
The charge of the cell membrane takes time to be modified. This time is dependent on the
membrane time constant which is related to the panel of ion channels and in particular the density of
voltage-gated sodium channels. Note that the composition of ion channels is different according to
the type of cell but also to the membrane region; some parts of the cell are more easily affected. The
chronaxie is a specific measure of this constant of time and represents the length of time necessary
to aim an activated state with a current equal to the double of the rheobase (Lapicque, 1930). The
rheobase being the current that is necessary to achieve the activation of the cell in an infinite time
(practically this activation arises around 300 ms). Comparable to a RC circuit from an electric point of
view, the nervous cell membrane has a time constant representing the product of its resistance (in
ohms) and its capacitance (in farads). In nerve cells the time constant of neural membrane has been
measured with TMS by Barker and colleagues to be around 150 µs (Barker et al., 1991). More
recently Peterchev and colleagues have used a controllable pulse parameter TMS (cTMS) device to
achieve the measurement of 196 µs for this time constant of motor cortex cells, and a chronaxie of
127 µs (Peterchev et al., 2013). The moderate difference in the measured time constant might be
due to the differences in materials (TMS system, coil, neuronavigation) and methods (orientation,
25
coil positioning, normalisation, and number of subjects) between the two studies. On the other hand,
the relation between the current change amplitude and the duration of the change regulates the
action of the electromotive force that makes the ions pass through the membrane – i.e. the higher is
the amplitude and/or the longer is the time, the more efficient is the stimulation. Thus the width of
the pulse wave is responsible for a more or less effective stimulation. Indeed, the increase in pulse
duration decreases the threshold of excitability that can be measured (Barker et al., 1991; Rothkegel
et al., 2010b). Consequently, the increase of the pulse duration would reduce the intensity of
stimulation for the determination of the motor threshold, to some extent (see figure 4). It also
reduces the pulse-to-pulse variability of the contralateral silent period for instance. The membrane
constant being variable between the different types of nervous cells, the adjustment of the pulse
width could constitute an interesting manner to stimulate different populations of cells selectively
(Peterchev et al., 2008). In spite of the importance of the current length, most commercially available
TMS systems only allow regulating the output intensity of the stimulator to modify the pulses – i.e.
one modifies the intensity of the current flowing through the coil wires and then the magnetic field
strength and the intensity of the induced current – and sometimes the choice between monophasic
and biphasic current.
Figure 4 from Rothkegel et al., 2010b: Motor threshold. Resting (RMT) and active (AMT) motor threshold for the longer
(simultaneous) pulse are significantly lower in terms of maximum stimulator output (i.e. capacitor voltage) compared to
the standard (single) pulse. Boxes represent all values between 25th and 75th percentile with a horizontal bar at the
position of the median. Upper and lower Whiskers represent the maximum and minimum values, respectively. ***
marks highly significant differences (p < 0.001, paired two-tailed t-tests).
26
An emerging type of generator has been introduced for the past few years, allowing more control on
the pulse parameters. The number, the polarity, the duration, and the amplitude of the pulse phases
can be adjusted. They are called controllable pulse parameter TMS (cTMS) and could allow for new
insights in clinical and research TMS (Petershev et al., 2008; Peterchev et al., 2011). They are capable
of the generation of high rate and high power unidirectional electric fields showing a near-triangular
pulse shape with less electric power consumption and coil heating for comparable or better
stimulation. The triangular shape of the pulse being known to be more efficient in the stimulation of
nervous tissue.
2.4 Different coil types and their incidences
The coils are constituted of wire loops and are equivalent to inductors. The amplitude of the
electromotive force induced by the system is dependent of the variation of current through the time
(in ampere by second) but also proportional to the inductance variable intrinsically related to the coil.
Thus the coil loop construction strongly influences the distribution and the power of the magnetic
field. The more the inductance of the coil is high the more the electromotive force is high. With any
coil architectures (flat multiple-layer disk coil, multiple-layer cylinder coil, or long single-layer cylinder
coil), the inductance of the coil is proportional to the permeability of the coil core and to the square
of the number of loop in the coil. The radius of the loop and the measurements of the coil section
will also influence the electromotive force, respectively increasingly and decreasingly. For the radius
of the loop, the situation in term of stimulation has to be nuanced, indeed a small loop produces
more intense magnetic field close to it but this induced magnetic field will decrease more drastically
with distance compared to a larger loop (Epstein et al., 1990). The heating and internal repulsive
forces in the wires are also some concerns of reduced size coils (Cohen and Cuffin, 1991). The shape
and size of the coil will of course define the incidence of the stimulation in the underlying nervous
tissue. Thus circular coils would produce a cylindrical magnetic field and big coils would produce a
more penetrating magnetic field for instance. The figure-of-8 coil represents the other common
shape of coil, merging two circular coils by one of their side in the same plane. The produced
magnetic field with this type of coil is really specific and shows an interaction between the magnetic
fields from the two circles of wires. Independently of tissue conductivity, the maximum of induced
current density is more focal compared to a single circular coil, and concentrated in the middle of the
two wings perpendicularly to the plane of the coil. It is the type of coil utilised in the two works
present in this report in order to induce transient stimulations on precise areas. Some variants of the
figure-of-8 coil have been developed showing an angle between the wings differing from 180 degrees
(flat plane). With this non-planar coil, the inclination between the two wings increases the depth of
27
penetration of the magnetic field concentrated at the bisection of the wing angle. Other types of
particular coils are also developed in order to achieve deeper stimulation (Roth et al., 2002; Zangen
et al., 2005; Carbunaru and Durand, 2001), but are up to now rather restricted to magnetic fields
experimentation.
2.5 Coil handling
With a given coil, the placement on the head surface is of course primordial in the
achievement of a proper stimulation. In order to induce the more accurate magnetic field, it is
generally required to apply the coil surface tangential to the curved surface of the head at the point
of contact. The magnetic field, being perpendicular to the coil surface, should be positioned such as
its path is the shortest to reach the target-area on the cortex. In this case, the amount of traversed
tissue is minimal and the reduced distance to the target allows a maximal incident electrical field in
the target zone. The output power of stimulation is sufficiently efficient with a limited intensity
output and a limited diffusion of the magnetic field to the adjacent areas. In case of a deep target,
the larger diffusion of the magnetic field at the surface layers remains an issue given that higher
intensities are needed for deeper stimulation. With some particular gyri configurations or according
to the adjacent regions and the targeted function, the magnetic beam could be oriented in a
different manner than tangentially to the head surface but rather in a position that follows the
inclination of a preferred gyrus or that avoids unwanted interaction with other regions. The circular
coil produces a circular electric filed, and the electric field of a figure-of-eight coil is concentrated in
the centre. The consideration of coil orientation is important with the latter type of coil. In addition
to the position of the coil on the head surface, the orientation of the coil, and consequently the
direction of magnetic field, influences also the efficiency of the stimulation. Indeed in the case of
figure-of-eight coils the orientation is matter of interest due to the current direction and its
interaction with the different cell layers. For instance, at the level of the primary motor cortex of the
hand, the best coil orientation (lowest threshold, greatest response) seems to be with the handle
directed postero-laterally in order to elicit a hand motor response. The coil is approximately
perpendicular to the central sulcus. In this case, the induced initial biphasic current in the coil
produces a more efficient stimulation if it passes in the antero-posterior rather than in the posteroanterior direction (Amassian et al., 1992; Brasil-Neto et al., 1992; Mills et al., 1992; Maccabee et al.,
1998; Kammer et al., 2001; see figure 5). Note that the current induced at the cortical level by the
initial wave of an antero-posterior biphasic pulse is opposed and goes in the postero-anterior
direction. Moreover, as developed above, the monophasic pulse currents and the initially biphasic
currents show inversed direction-efficiency ratio. It is noteworthy that compared to biphasic, the
28
monophasic pulses are more sensitive to orientation in both single pulse and rTMS paradigms due to
their unidirectional current. Though, for a similar induced current, different muscle tracts have
shown diverse optimal orientations (Maskill et al., 1991; Terao et al., 1994, 2000; Guggisberg et al.,
2001). In fact the best orientation could also depend on the muscle tract targeted in a single coil
position (Pascual-Leone et al., 1994). The motor manifestation modulated by coil orientation is
however related to modification in the cerebral signal when it is recorded online – e.g. some missing
components for the non optimal orientations (Bonato et al., 2006). This sensibility to current
orientation is related to the fact that neurons are more easily excited by a current flowing in parallel
to their fibres as showed by Rushton on peripheral nerves (Rushton, 1927). Based on the different
studies of the response to motor stimulation with a current perpendicular to the gyrus and the
assumption of a preferential activation of nerve cells with tangential currents, plus the results form
time conduction studies, the cell population triggered in hand motor cortex has been identified as
the interneurons with afferent fibres to the pyramidal cells of the hand motor cortex (indirect wave).
A similar modulated response to stimulation according to the induced current orientation is expected
to occur for the other areas in the cortex; since the fibres of the local population of nerve cells are
organised with predominant orientations. However, according to the local configurations of the
fibres of the different populations of cells, the stimulation could be elicited directly on the pyramidal
Figure 5 from Kammer et al. 2001: Variation in motor thresholds of 8 subjects. Two different TMS systems were used,
the Dantec and the Magstim. On the abscissa the different stimulation configurations are given. Waveforms were
biphasic or monophasic, orientations of the induced current were antero-posterior (a-p) or postero-anterior (p-a).
Threshold values x0 are given as a percentage of the maximal output of the stimulator used. Each symbol represents
one subject. In conclusion, monophasic or biphasic stimulations induce their maximal effects on the motor cortex with
opposite direction of the initial current.
29
cells, by interneurons or both. The resulting effects would be diverse according to the cell group
targeted. When recorded online, the cerebral response is modulated according to the current
orientation in the area stimulated with weaker brain manifestations for non optimal orientations
(Thomson et al., 2012). These elements argue for a specific current orientation for each target-area
according to their composition in term of cell population and fibre orientation. The orientation of the
induced electric field has to be chosen to impact the target gyrus and to avoid acting on the
neighbouring gyri if a specific effect is desired. Numerous studies have looked at these phenomena in
the motor cortex however more investigations and reports are needed to have a better idea of the
incidence of current orientation in any areas of the cortex.
With adaptation of the pulse waveform, amplitude and width, the type of the coil and the
orientation of the induced current according to the orientation of the axon fibres, it might be
possible to selectively target a population of neurons in a limited area of the cortex (Di Lazzaro et al.,
2004). In addition as developed below the type of stimulation (single pulse, train of pulses or
repeated according to diverse protocols) allows to control the effect of the stimulation.
3. Protocols of stimulation
The different effects of TMS are related to many parameters and maybe more particularly to
the sequence of pulse release. Indeed, TMS experimenter may attempt to add some transitory noise
with isolated pulses or conversely to induce long-term effects and cortical excitability changes using
repetitive TMS. The different protocols of stimulation developed and assessed along the years are
shortly described in the following paragraphs.
3.1. Single pulse TMS
In order to induce transient noise in an area or a network, sporadic pulses are released as
single pulse or in train of pulses to a specific cortical target. The isolated stimulation of a single or a
sequence of pulses can be varied in term of intensity, waveform, and orientation of the coil. Freely
released stimulation can be used in MEP studies or in combination with neuroimaging. In cognitive
experimentation isolated pulses are released in a time-controlled manner occurring at a precise predefined moment during a task. The paradigm of stimulation is then termed event-related stimulation.
The area and the timing defined by the hypothesis of the study are chosen to maximize the
physiological effect of the TMS in the task. Different timing of pulse launch can also be assessed in
order to test the chronology of implication of the target-area. Such isolated pulses disturb or add
noise transiently in the neuronal system via the target-area as an entry (Harris et al., 2008). A
30
common protocol consists in isolated train of pulses released in a controlled manner in a specific
time window related to stimulus onset. This generates a disturbance induced by several pulses
(between 3 and 5) at high frequency (between 20 and 10 Hz) covering some hundreds of milliseconds.
While isolated event-related pulse aims at assessing the function of an area at an exact time, the
train of pulses aims at assessing the function of an area during a more or less extended short period
of the order of the fraction of a second. The delay between the events of stimulation, isolated pulses
or trains, defines the occurrence of long-term effects. A delay of 3 seconds was observed to avoid
any additive effect between the isolated pulses (Rothwell et al., 1999). However, the 3 second limit
seems to have been underestimated (e.g. Ikeguchi et al., 2005).
3.2. Repetitive TMS (rTMS)
Conversely to the desired short time action of the isolated stimulation, the rTMS protocol
aims at modifying the brain function on a time outlasting the stimulation period. The supposed main
effect of rTMS is the immediate modification of synaptic connections. The nature and strength of the
long-term effects of rTMS are highly dependent on the various parameters of stimulation, and
numerous exceptions showed the need of more studies to clarify the relation between the different
aspects of the stimulation protocol and the output effects. Commonly the continuous low frequency
rTMS of hundreds of pulses lead to an inhibition of the targeted area and a lowering of its excitability
while a continuous high frequency rTMS protocol leads to a facilitation effect with an elevation of
excitability (Fitzgerald et al., 2006; Pascual-Leone et al., 1994; Maeda et al., 2000; Chen et al., 1997;
Muellbacher et al., 2000; Cao et al., 2013). Classically the low frequency protocols use stimulation at
1 Hz or below and high frequency protocols use stimulation at 5 Hz or upper. Due to safety
consideration, in the case of high frequency stimulation, breaks of several seconds without
stimulation are generally introduced during the protocol dividing the sequence of stimulation into
blocks (see Table 5 from Rossi et al., 2009). The total number of pulses in rTMS protocols varies
between 500 and 2000 while the intensity of stimulation represents 90 to 130 % of the motor
threshold.
3.3. Patterned and Theta burst stimulation (TBS)
Newly introduced in the middle of the 2000’s (Huang et al. 2005), TBS has appeared as a
quick and efficient method to induce long lasting effect on excitability and functioning of a cortical
area. TBS is a patterned alternative of repetitive TMS. It is constituted of bursts of high frequency
pulses (classically 3 pulses at 50Hz) released on a theta rhythm (5Hz). Hence the burst onsets are
31
separated by 200 ms. From the first attempt of TBS intensity of the stimulation was kept relatively
low – generally at 80% of the active MT. Some variants exist with bursts at 30Hz released every
166ms and are qualified as modified TBS (Nyffeler et al., 2006; Goldsworthy et al., 2012). These
protocols, classic or modified, represent constant TBS (cTBS) and last generally less than a minute
with 600 pulses in 200 bursts. However, another protocol introducing breaks of 8 seconds every 10
seconds is the so-called intermittent TBS (iTBS). The effect of the two types of protocol were tested
and while cTBS showed inhibitory effects on MEP, iTBS showed facilitation of MEP when stimulate
the corresponding motor cortex (Huang et al., 2005 see figure 6; Zafar et al., 2008). Noteworthy, the
breaks responsible for the facilitation effect of iTBS showed similar effect in a repetitive TMS block
protocol (Rothkegel et al., 2010a). Effects of the different TBS protocols on higher cognitive function
are less well elucidated. While the observed effects of TBS last as long as or longer than after classic
rTMS protocol, TBS has the advantage of being more comfortable for the participants than most
rTMS protocols. The TBS protocols are relatively faster and performed at limited intensity. Also TBS
has been associated with a reduced risk of mild and severe adverse events compared to the
conventional rTMS (Oberman et al., 2011). All together these advantages lead to an increasing use of
TBS.
Figure 6 from Huang et al., 2005: A. In the intermittent theta burst stimulation pattern (iTBS), a 2 sec train of TBS is
repeated every 10 sec for a total of 190 sec (600 pulses). In the intermediate theta burst stimulation paradigm (imTBS),
a 5 sec train of TBS is repeated every 15 sec for a total of 110 sec (600 pulses). In the continuous theta burst stimulation
paradigm (cTBS), a 40 sec train of uninterrupted TBS is given (600 pulses).
B. Time course of changes in MEP amplitude following conditioning with iTBS (closed up triangle), cTBS (closed down
triangle), or imTBS (open circle). There was a significant effect of pattern of stimulation on change in MEP size following
stimulation [F(2,16) = 20.32, p 0.001], with significant post hoc differences between each pattern of stimulation. There
was a significant facilitation of MEP size following iTBS lasting for about 15 min, and a significant reduction of MEP size
following cTBS lasting for nearly 60 min. imTBS produced no significant changes in MEP size.
32
3.4. Paired TMS
Paired TMS may assess the connectivity between two areas using a double isolated
stimulation. One area is targeted by a first coil and the second one by a second coil. The likely
influence of the first stimulation is assessed in the response change to the second one – often
directed towards the primary motor cortex. In the case of a proved efferent connectivity of the area
one to the motor cortex as secondary area the recorded MEP will be modified. In order to observe a
change in the response, a particular delay – generally around dozen of milliseconds – between the
first and the second has to be respected. The two stimulations could be qualified as mutual eventrelated stimulation. The two coils as to be connected to the same machine or two machines have to
be temporally linked together or by a third element – i.e. a computer via TTL cables.
4. Interests of the technique
TMS is a non invasive tool used to achieve electrical stimulation of neural tissue in the
cerebral cortex. Within certain conditions of utilisation, TMS is used in healthy participants or
patients as a convenient manner to induce cortical stimulations for different kinds of paradigms. The
main utility of TMS at the beginning was to assess neural excitability of the central nervous system.
Indeed its simplicity and the fact that it is not invasive and relatively comfortable for the subject
made TMS a good method to study neural excitability. TMS is also a “causal functional imagery
technique” without direct image output of the brain structures or activities as collected with the
other classical imagery techniques. The “imagery” of the TMS is indirectly deduced from the causal
link established between an area of nervous tissue which is stimulated and a change in a studied
function. Withal, the effects collected after the stimulation of a specific area allow reconstructing a
causal cartography – the implication of an area in a motor or sensory function, in a functional
network, or in a task. In this sense the TMS is a tool of choice to validate the causal implication of a
particular area or network in a specific function. Repetitive TMS is also used for its potential longterm effects with neurophysiological and excitability changes. Hence rTMS notably showed ability to
change, enhance or inhibit the excitability of a stimulated area or areas related to the stimulated
area. It is of particular interest for treatment of some neurological or psychiatric diseases.
The request in PubMed search engine of the National Center for Biotechnology Information with the
terms transcranial magnetic stimulation reaches 9939 results, with 1205 only for the year 2013.
4.1. Use of TMS as a research tool
33
4.1.1. Measuring neural excitability
The TMS is a tool of choice to measure the cortical and nervous excitability, because it
permits to induce electrical stimulation of the cerebral cortex in a reliable manner and without injury.
The power of output and the interval between pulses can be easily set to reach the sought
stimulation parameters. When applied over the motor cortex, the muscular responses to the
stimulations may be observed and recorded in term of amplitude, duration, or latency. Also “silent
periods” are of interest. The TMS can be applied over the cortex but also over the spinal cord or
peripheral nerves and thus can be used to measure time of nervous conduction at different levels of
the motor system. All these measures are particularly interesting in patient populations with motor
disease. Hence TMS constitutes a way of studying the motor functions in different conditions or
groups. When TMS is applied over the occipital cortex, phosphenes can be induced. However, in
contrast to the motor system, the perceived phosphenes are not objectively measurable. They can
only be reported by the participant and thus they are subject to his own perception.
As direct measurable variables, the motor manifestation and phosphenes perception constitute the
two manners to evaluate individual threshold of cortical excitability to magnetic stimulation,
respectively motor threshold (MT) or phosphene threshold (PT). These thresholds represent the
minimal power of the magnetic field necessary to collect a response from the participant. It is
particularly important to apply stimulations adapted to the excitability of each participant for further
comparisons, and to avoid an overstimulation of the cortex which could be deleterious for the
participant health (see Rossi et al., 2009 for guidelines). Commonly the individual cortical excitability
is evaluated with the motor response of the fingers, and more precisely of the adductor pollicis
muscle, while the TMS coil is placed over the contralateral motor cortex of this group of muscles. In
this case the stimulation may produce a twitch of the muscle. The threshold is reached when 5 out of
10 stimulations (50 %) induce a motor response. The response is considered valid when 50µV current
is measured with an EMG on relaxed muscles – generally the abductor pollicis brevis muscle. It is
named the resting motor threshold (Rossini et al., 1994). Although it is the main method currently in
use in the evaluation of the motor threshold, many studies evaluate the motor threshold without
EMG measurement by considering a twitch of the fingers by visual inspection as a motor response.
This is sometimes named clinical motor threshold. The principle of 50% of responses remains the
same but the visual inspection has a tendency to overestimate the needed output power compared
with an EMG-measured threshold. Like motor responses, phosphenes can be used to evaluate an
individual phosphene threshold (PT). It is particularly interesting to determine the excitability of the
occipital lobes when the following stimulation is supposed to be in this area. When TMS is applied
over visual areas, the phosphene threshold represents the power output necessary to induce a
34
phosphene perception in the contralateral hemifield. However the phosphene threshold measure is
less precise than the motor threshold (Afra et al., 1998; Aurora et al., 1998), due to the absence of
external detection system, and is thus forsaken compared to the motor threshold measure in the
case of generic neural excitability studies.
The purpose of an adaptation of the power of stimulation to these thresholds is the induction of
stimulations that are comparable and adapted to the individual cortical excitability. It is particularly
important in group studies. However this adaptation to the excitability threshold measured in one
area raises the controversy about the absolute consistency between the different head regions and
cortical areas. Distance from the surface, cell composition or orientation and different other reasons
lead to doubt the relation between different areas in term of response to TMS (Kozel et al., 2000;
McConnell et al., 2001; Stewart et al., 2001; Boroojerdi et al., 2002; Gerwig et al., 2003; Deblieck et
al., 2008). Interestingly, Kähkonen and colleagues demonstrated that the stimulation of the motor
and the premotor cortices produces different amplitudes of evoked response in EEG but these
responses are positively correlated (Kähkonen et al., 2005).
The measure of motor excitability constitutes withal an index of cortical excitability. Thus it is often
used in the study of motor pathologies. Huntington’s disease (Lorenzano et al., 2006), Parkinson’s
disease (Vacherot et al., 2010 or Derejko et al., 2013), or multiple and amyotrophic lateral sclerosis
syndrome (Nardone et al., 2005 or Bae et al., 2013 in ALS; Mori et al., 2013 or White et al., 2013) are
some of the diseases studied with TMS on a neuromotor level. TMS is also of interest in the study of
other pathologies not directly related to motor dysfunction but to cortex excitability as Alzheimer’s
disease (Alagona et al., 2001) or migraine (Brigo et al., 2013). The measure of excitability by TMS can
also be used to evaluate the changes induced by a previous rTMS protocol (e.g. on the same area
Muellbacher et al., 2000; or the contralateral area Schambra et al., 2003) or simply in different
conditions (e.g. Giovannelli et al., 2013). In addition to the determination of individual threshold of
excitability, TMS is used in studies about excitability to measure the amplitude and time of muscle
contraction, the time conduction, the intracortical inhibition or the central and peripheral silent
periods, in response to simple or repeated stimulations, at rest or during voluntary contraction, and
in different groups (healthy, patient, particular populations) or in different conditions (medication,
training, rTMS). The amplitude and variability of contraction in response to comparable stimulations
indicate an alteration of the motor command circuitry. The time conduction measure is obtained by
comparing the delay between stimulation and muscle response while the coil is placed over the
motor cortex, the spinal cervical fovea or the peripheral nerves and informs on the ability of the
motor system to transport neural information; e.g. an increase in time conduction could be indicative
of a demyelinisation. Intracortical inhibition is measured with paired pulses TMS and is related to
35
inhibitory interneuron integrity. Interhemispheric or interarea inhibition could also be measured with
paired pulse stimulation conveyed in this case through a double coil protocol – e.g. between two
homologous areas. Also, excitability measures performed during different times allow assessing the
plasticity of cortical areas, for instance in Parkinson patients (Rothwell, 2007) or in sleep apnea
syndrome (Das et al., 2013).
Based on the measure of motor cortex excitability, the evaluation of connectivity has been
developed using two magnetic stimulators to generate paired-pulses. A primary single pulse
(conditioning pulse) launched over an area influences the event manifested after a second single
pulse (conditioned pulse) to another area. Hence the impact of the conditioning pulse could change
the signal collected after the conditioned pulse. The second target-area is usually the motor cortex
and the signal measured is the MEP. When the two areas are functionally connected, the effect of
the second pulse is likely to be modified. The connectivity is defined in terms of direction (excitatory
or inhibitory), strength and timing with the favourable inter-pulse interval. One of the important
domains of application of this specific measure is the inhibitory inter-hemispheric connectivity – i.e.
the inhibitory fibres of the corpus callosum. In this case the connectivity between the two
homologous primary motor cortices of the hands is engaged by placing the coils over both areas as
introduced by Febert and colleagues in 1992 (Ferbert et al., 1992). The strength and the duration of
the interaction are dependent on the intensity of the conditioning stimulation (Netz et al., 1995). The
connectivity between the two primary motor cortices could be measured in different conditions; for
instance the inhibition from right to left was reported to be decreased during speech production by
Kano and colleagues (Kano et al., 2012). Interhemispheric connectivity could also be approached
using a single coil and by measuring the effect on ipsilateral muscles (Meyer et al., 1995) or in the
sensibility of the ipsilateral limb (Seyal et al., 1995). The double-coil approach has been used in the
visual modalities as well. For instance Pascua-Leone and Walsh demonstrated the feedback action of
V5 on V1 in the awareness of visual motion (Pascual-Leone and Walsh, 2001).
4.1.2. Modification of excitability and long term effects
An important and more recent application of TMS is the induction of relatively prolonged
modifications of brain functioning. These changes are induced by specific protocols of TMS that
implement repetitive stimulations (including rhythmic and patterned stimulations). An isolated pulse
is able to activate a group of cortical cells of the brain and when repeated in trains of pulses it can
modify the cortical excitability (Chen et al., 1997; Muellbacher et al., 2000), the metabolism (Nahas
et al., 2001; Thiel et al., 2006; Cao et al., 2013), the plasticity of brain structures (Baümer et al., 2003;
36
Zieman, 2004), and on a larger scale the behaviour. These effects outlast the stimulation time and
could also occur remotely to connected areas by efferent fibre connections (Baümer et al., 2003;
Thiel et al., 2006; or in animal models by Valero-Cabre et al., 2005). The type of effect is dependent
on the stimulation frequency, and its strength and duration are dependent on the parameters of
stimulation and the number of pulses released – not always clearly defined as developed hereafter.
The waveform of the pulse and orientation of the coil may selectively modify the impact of the TMS
on different populations of neural cells of a given area. This relationship is tricky to characterize but it
is obvious that these parameters are potentially important for the output effect of rTMS. The main
mechanism responsible for persisting effects of TMS is often related to plastic changes in synaptic
functioning. The factor defining the synaptic plasticity and its “direction”, long term potentiation
(LTP) or long term depression (LTD), is conveyed by the relation between cell activations and time.
Indeed the synapse could be strengthened or weakened between the pre- and post-elements, as the
rTMS could potentiate or depreciate the cortical excitability. Both phenomena are linked to the
frequency of stimulations. LTP and LTD involve high and low frequency stimulation, respectively,
even if in fact these two categories are mediated by multiple molecules and cellular mechanisms and
are likely to be different according to the cell population considered and the characteristics of the
individual (Malenka and Bear, 2004). Similarly rTMS effects have been classified in two categories of
excitatory or inhibitory effects induced by high or low frequency protocols of stimulation respectively
(Fitzgerald et al., 2006; Pascual-Leone et al., 1994; Maeda et al., 2000; Chen et al., 1997; Muellbacher
et al., 2000; Cao et al., 2013). In the case of rTMS over the motor cortex, the couples highfrequency/activation and low-frequency/inhibition are rather robust (Fitzgerald et al., 2006).
Nevertheless the frequency-dependent effect of rTMS seems to be not so straightforward and
contrary effects were observed in other cortical areas. As notified by Pell and colleagues (Pell et al.,
2011), other factors than frequency alone may influence the effect of rTMS – e.g. patterns of pulse
release, nature of pulses, train duration or intensity of stimulation. Also diverse neurophysiological
mechanisms, other than LTP/LTD, are possibly responsible for the long-term changes observed after
rTMS. There is a lack of pharmacological studies dismantling the effects of rTMS, however the few
existing studies indeed argue in favour of synaptic plasticity (Pell et al., 2011). The different scenarios
of the TMS-induced changes are also resumed by Wagner and colleagues (Wagner et al., 2009).
4.1.3. Measuring indirect effect on behaviour
By altering the functioning of areas and circuits of the brain, TMS is also potentially able to
modify behaviour. Contrary to motor and sensitive modalities that can be studied through direct
responses, the changes in cognitive function after TMS can only be determined indirectly through
37
specific and well controlled tasks. These tasks have to be related to the investigated function of the
area or the circuit targeted by TMS. TMS is widely used in neuropsychological research given the
potential interest of determining causal relations between brain areas and functions without having
to refer to brain lesion studies. In addition to the localisation of a function, TMS in event-related
paradigm allows determining the timing of a given cognitive process. The pulses are considered to
add immediate neural noise in the area and disturb the brain signal (Harris et al., 2008) while the
repetition of pulses is capable to induce consecutive fatigue of neural populations or to change their
excitability for a variable duration. Repetitive TMS can enhance or inhibit the excitability depending
on different parameters of stimulation and the stimulated cortical area as developed in the previous
paragraph. Accordingly, the stimulation protocol has to be adapted according to task and the desired
modification. The induced perturbation is expected to be either concomitant to the stimulation when
the TMS is applied as in the case of event-related pulses, or to have relative long-term residual
effects after a repetitive stimulation protocol. Consequently, the effects of the stimulation have to
occur at the particular time when the task is effectively tested. In other words, the task has to be
adapted to the stimulation protocol and to test the participant abilities when the TMS is expected to
be the most effective. The behavioural changes are generally moderate so the task has to be fine
tuned to detect behavioural changes. The alteration of behaviour asserts an important role of the
targeted structures in the proper realisation of the considered task. The behavioural changes could
concern the accuracy and/or the reaction time and/or the discrimination.
Many modalities have been studied in event-related or repetitive TMS protocols. For example, the
alteration by rTMS of diverse areas (e.g. anterior intraparietal sulcus, prefrontal cortex, anterior
temporal lobe or inferior parietal lobule) induced the modification of object processing and
recognition (Buelte et al., 2008; Viggiano et al., 2008; Ishibashi et al., 2011). Disruption of feature
binding and spatial attention processes have been achieved by inhibitory rTMS over the right parietal
cortex, leading to a facilitation of visual perception (Oliveri et al., 2010). Conversely high frequency
rTMS induced a pseudo neglect by facilitating orientation to the contralateral hemifield (Kim et al.,
2005); otherwise the high frequency rTMS over the left dorsolateral prefrontal cortex modified
attentional control and improved inhibition (Hwang et al., 2010). Memory aspects are also studied
using TMS. Working memory (Mottaghy et al., 2003) and prospective memory (Bisiacchi et al., 2011)
were modified offline by rTMS or online by event-related TMS. Also language functions have been
studied a lot using TMS paradigms, starting with the “historical” TMS-induced speech arrest by
Pascual-Leone and colleagues in 1991 (Pascual-Leone et al., 1991). The potential of TMS to define
language lateralisation and dominance was then investigated in view of a potential clinical utility
(Jennum et al., 1994; Michelucci et al., 1994). More recently, TMS has been adopted as a lesion38
based method for language mapping in patients (Tarapore et al., 2013). Thanks to TMS experiments,
fine aspects of language processing could have been attributed to different brain regions. TMS
disentangled roles of anterior and posterior left IFC implied in semantic or phonologic aspects
respectively (Gough et al., 2005). Unexpected premotor cortex was found to be implicated in action
language understanding (Willems et al., 2011). Emotion processing is also an important cognitive
function investigated with TMS. Most of these studies explored the recognition of different kinds
and/or intensities of emotions in various modalities. Processing of emotion in prosody has been
localised in both right anterior superior temporal gyrus and fronto-parietal operculum more
particularly for the withdrawal emotions compared to the approach emotions in an event-related
TMS experiment (Hoekert et al., 2008). Bilateral cooperation in temporoparietal regions was
identified in emotional word reading in the second study of this report (Rochas et al., 2014). The
prefrontal cortex has been implied in emotional expression recognition (van Honk et al., 2002;
Rochas et al., 2013) and its priming (Mattavelli et al., 2011). Conversely the influence of the
emotional state on cortical excitability has also been demonstrated by TMS, that is to say a negative
emotional context induced an increase in MEP amplitudes thus showing that emotional states might
have a direct impact on motor system (Baumert et al., 2011). Concomitant fast rTMS over left
dorsolateral prefrontal cortex disturbed the perceived-controllability effect on the emotional
dimension of pain experience (Borckardt et al., 2011). Pain perception itself has been related to
prefrontal cortex in TMS experiments (Barckardt et al., 2009; Krummenacher et al., 2010). TMS has
even been used to modify the behaviour of drug addicted patients; high frequency rTMS over the
right DLPFC transiently reduced the craving sensation towards cocaine addiction (Camprodon et al.,
2007).
In principle, any cortical regions reachable by TMS could be subject to investigation with a
corresponding appropriate behavioural task. TMS constitutes a powerful tool in the study of short to
long term plasticity by induction of virtual lesion and, in the construction of functional causality maps
between areas and functions. In addition event-related TMS investigation allows getting information
on time of involvement of the targeted structure in the studied function. However, the moderate
effects of TMS are often tricky to interpret and might engage several aspects of a function or several
functions mixed statistically through interactions. The modification of the functioning of one area is
sometimes difficult to disentangle from the action at the network level. Thus the coupling of the TMS
with another brain imagery technique could help in the understanding of the TMS effects on the
function of the underlying neural network.
4.1.4. Performance enhancement
39
Taking into account that TMS is a method to induce a transient and virtual lesion, the
improvement of skills by TMS was not initially obvious. Nevertheless, as developed above, the TMS is
more generally a tool for non invasive stimulation of the cortex and can in this way modify local brain
excitability as well as motor performance, perception and higher levels of cognition. Related to these
effects on cognition and behaviour, a potential enhancement of skills may be achievable. TMS is used
with success in rehabilitation of lost functions, and thereby could theoretically be used in the
enhancement of functions at a higher level, above “normality”, in a healthy individual. Hence, recent
interest in the potential of TMS in terms of amelioration of performances has increased (Luber and
Lisanby, 2013; Brem et al., 2013). Possibly, the improvement could be induced in a direct or an
indirect way. The direct way consists in the modification by TMS of the function of a cortical area or a
network in a way that it reacts more efficiently – i.e. more accurately and/or more rapidly – to a
specific task. The direct way implies an enhancing action on structures directly implied in the task.
The indirect way of performance enhancement through TMS would be mediated by the annihilation
of a physiologic but “negative” or limitative action of an area towards the network of interest in the
given task. The more probable case would be the TMS-induced inhibition of a cortical actor that is
naturally disturbing or competing with the network involved in the task, consequently the task will be
executed more efficiently. This case is well represented in the interhemispheric inhibition found
between homologous areas. The neuroenhancement by disruption follows an addition-bysubtraction mechanism. The inhibition of an area may be achieved by pulse-related and transient
noise or more prolonged by rTMS – i.e. classically low constant frequency or patterned stimulations.
As in any TMS paradigm, the dichotomy between short and long-term effects by event-related or
rTMS is again of importance in this issue. The enhancement of performance would be either induced
(a) live by time-locked TMS pulses during a task or (b) for a period following the rTMS. These two
approaches lead to different views. In the first case (a) the performance is augmented due to a
virtual lesion or noise input to the brain, and would disappear when the TMS is off. This on-off ability
may have an interest as the enhancement of one skill could bring on the other hand a detrimental
impact on others as developed below. A precisely time-locked and transitory potentiation of one step
of processing by a single pulse or train of pulses on an area would let the residual brain intact for
other steps of a function and for other functions. In event-related TMS, a time-tuned enhancement
could be possible. According to the review of Luber and Lisanby, the online TMS enhancement is
predominant compared to offline enhancement (see Table 1 in Luber and Lisanby, 2013). Luber and
Lisanby or Brem and colleagues mentioned also an amazing mechanism called stochastic resonance
(Luber and Lisanby, 2013; Brem et al., 2013). The mechanism has been introduced in TMS domain by
Miniussi and colleagues (Miniussi et al., 2010) and was derived from Gammaitoni’s work
(Gammaitoni et al., 1998). According to this principle, the addition of a gentle amount of noise by
40
TMS (intensity near or under excitability threshold) in a specific structure could push a subthreshold
physiological signal – e.g. weak sensory input, to be detected and pass beyond the structure to reach
further steps of processing. The summation of the natural information and the moderate noise
activity generated by TMS would reach the physiologic threshold for attention, detectability, motor
command or any further support of this hybrid signal. The amount of noise brought by TMS has to be
within a precise power window; a too low level of noise would not help enough that the information
can reach threshold of processing and a too high level of noise would reduce too much signal-tonoise ratio destroying the initial information. In addition, the stimulation has to be perfectly timelocked in the targeted region with the natural processing and it is a specific case of short term eventrelated enhancement. A third mechanism of online neuroenhancement is represented by the
inductions of oscillatory activity in the brain. This mechanism relies on the principle that certain
oscillatory rhythms induced in certain regions would favour some functions (or inhibit some others).
The entrainment of specific oscillations by TMS has been used to act on rather short-term periods
(Romei et al, 2011; Thut et al., 2011b), supposing no outlasting plastic effect and a return to basic
oscillatory state after the end of the stimulation. In the second case (b) though, the effect is expected
to last decades of minutes after the stimulation – i.e. with repetitive stimulation. LTP-like processing
can be involved and excitability would be modified at the impact point but also likely into the
downstream structures. The rTMS treatment is more prone to play with plasticity mechanisms and
therefore engages more complex and deep modifications. A third aspect of performance
enhancement by TMS may be conveyed by the side effects of TMS rather than by its main action of
stimulating the neural cells. In fact various sensory aspects of the stimulation in term of sound
(audition), vibration (somatosensoriel), excitation of peripheral nerves and muscles could by
themselves arouse attention of the participant and trigger unspecific ameliorative effects. Indeed
some studies found better performance in the task for all the conditions tested, comprising the
control conditions (Pascual-Leone et al., 1992; Campana et al., 2003; Drager et al., 2004). In sensoryrelated tasks, the unspecific TMS effect acted as enhancer in a phenomenon called intersensory
facilitation. This emphasises the critical importance of a good experimental design with a valid
control condition to disentangle the obtained effects which will be discussed in this manuscript.
4.1.5. Tracking of the TMS impact with brain imagery
Apart from motor or sensitive manifestation and behavioural changes, the effect of TMS may
be also monitored by means of brain imagery techniques. Indeed the combined measure of the brain
activity permits to determine changes not only in the direct target-area but also in remote areas
through the connected network. In case of rTMS, the changes could be reflected even in structures
41
that are not directly connected to the target-area, due to more long-term plasticity mechanisms. In
correlation with eventual measurable behavioural changes, the monitoring of brain activity helps to
explain the repercussion of TMS on the performance of the participant. It also allows a better
understanding of some of the paradoxical effects of TMS – i.e. unexpected orthogonal effects for
similar sites or types of stimulation, or absence of effects.
One imaging method to study TMS effects is the fMRI. The technique allows the acquisition of brain
activity image through blood oxygen level dependent (BOLD) signal without a priori hypotheses of a
particular location. It allows also looking at networks as well as anatomic and functional connectivity.
Thereby functional connectivity MRI constitutes a valid tool to be associated with TMS (Fox et al.,
2012). The image acquisition session may be online – i.e. during the TMS procedure with the TMS coil
in the scanner tube, to monitor the effect of TMS pulses or during a task in an event-related TMS
paradigm (for a review see Ruff et al., 2009). For instance suprathreshold TMS induced BOLD
responses have been demonstrated in the area corresponding to the targeted finger in the primary
motor cortex while subthreshold stimulation did not induce any signal detected in the MRI, as well as
during the suprathreshold stimulation of the premotor area (Baudewig et al., 2001). More recently,
online TMS-fMRI gave evidence of a top-down modulation from FEF on retinotopic visual cortex (Ruff
et al., 2006). The acquisition may also be directly consecutive to the TMS session; the participant is
placed in the MRI scanner following an rTMS protocol.
The positron emission tomography scanner (or PET scan) collects images from radioactive signal. Its
principle is based on the injection of invasive radioactive tracers. The electromagnetic signal from
TMS does not interfere with the radioactive signal of PET. An early combined PET-TMS study has
already shown positive correlations between the cerebral blood flow and the number of pulses and
remote effects of repeated TMS pulses from the FEF to visual areas (Paus et al., 1997). Remote
effects were depicted in PET after the stimulation of the hand primary motor cortex (Fox et al., 1997)
or the left Broca area (Thiel et al., 2006; see figure 7). Offline PET acquisition is of course also
envisageable. Due to its limited temporal resolution that depends on the half-life of the used tracer,
this technique will lead to a picture of compound TMS effects that are averaged over time.
42
Figure 7 from Thiel et al., 2006: Illustration of the effect of 4Hz rTMS on left IFG. (A) Images show IFG activation during
simple verb generation. (B) Images clearly show the decreased activity on the left (blue arrow) and increased activity on
the right side (yellow arrow) during rTMS interference. Online combination of TMS with neuroimaging allows for the
visualisation of remote effects of the stimulation.
Alternatively, EEG is a method to measure brain activity changes with the advantage of high temporal
resolution allowing direct tracing of TMS effects in cortical networks. Recent development of
algorithms and methodologies allows for an accurate spatial resolution by application of source
localisation algorithms on high-density EEG recordings (Michel et al., 2004). Therefore, the EEG
constitutes a perfect complement to TMS thanks to its accurate time resolution and the compatibility
of the two systems (for review Miniussi and Thut, 2009). The characteristics of EEG are especially
advantageous in online event-related paradigms where they permit to track and isolate the different
steps of the TMS disturbance and the modified neural processing. This can go from the early time
after a pulse to longer or further periods, or even before pulses. Recent improvements in hardware
and processing allow studying the EEG signals during single-pulse or repetitive TMS (Thut et al., 2005;
Rogasch and Fitzgerald, 2013). The information collected from EEG may be rather global. For instance,
Kähkönen and colleagues measured global mean-field amplitudes that reflect overall cortical activity
comparing different intensities and shapes of pulses, and different areas (Kähkönen et al., 2005).
Some of their results permitted to identify a weaker impact of comparable pulses on premotor
compared to primary motor area. Bonato and colleagues used EEG to monitor online TMS effects of
different orientations of the coil and different inclinations over the left primary motor cortex (Bonato
et al., 2006). Some of the inclinations were supposed to act as sham stimulation. Their results
invalidated the “uneffectiveness” of sham stimulation with the coil inclined at 45° or 135° over the
43
scalp and showed rather ERP induction even without motor manifestation with the coil at 135°.
Similarly, Ortu and colleagues found alterations in EEG signal after cTBS of left M1 while no change
was observed in EMG of voluntary movement (Ortu et al., 2009). This example shows an advantage
and the possible higher sensibility of imaging technique in the monitoring of effects compared to
motor (in the present case), sensory or cognitive outputs (Rossi et al., 2000; Hansenne et al., 2004;
Holler et al., 2006). Concurrent TMS-EEG studies also allow for the investigation of pre-state
influences on TMS pulse effects, especially in term of frequency band analysis (Schulz et al., 2013;
Zanto et al., 2014). Of course the compatibility of EEG and TMS online does not exclude the use of
offline EEG recordings after an rTMS paradigm. For instance, offline low frequency rTMS over the
striate cortex decreased the amplitude of CI, an early evoked potential component in visual
processing (Shutter and Van Honk, 2003). The high temporal resolution of the EEG also allows
studying the TMS effects on brain rhythms or, inversely, the influence of brain rhythms on TMS
effects (for review Thut and Miniussi, 2009). While spontaneous brain rhythms influence the cortical
excitability and thereby the impact of TMS (Romei et al., 2008; Sauseng et al., 2009), more
interestingly the application of rTMS could also cause rhythmic entrainment (Thut et al., 2011a).
4.2. Use of TMS as a clinical tool
Naturally a noninvasive tool for focal neurostimulation brings potential advantages in clinical
practice. In addition to fundamental research, TMS has become a standard clinical tool mainly for the
diagnosis of motor pathologies and constitutes a good approach to define the specific symptoms of
such diseases by measurement of abnormal evoked responses. Long lasting effects of TMS represent
also a possible treatment in certain cases of psychiatric and neurologic diseases.
In diagnostic use, diverse measures of motor activity under influence of TMS can be used as
pathological markers and inform about the nature or the gravity of these diseases (Groppa et al.,
2012). Different variables may be considered in the diagnostic of motor system disease: the cortical
motor threshold of excitability, the amplitude and delay of the TMS-induced MEP, the central motor
conduction time (the time taken by the nervous flux to go from the cortex to the motoneuron), the
cortical silent period duration (the time of muscular inactivity after TMS-induced MEP), the
transcallosal inhibition and various possible ratios between these measures. These different variables
are susceptible to be altered in different cases of pathologies touching the cortical excitability, the
corticospinal tract or the innervations of muscles, the myelinisation of nerve fibres (Lee et al., 2003;
Jhunjhunwalaet al., 2013; Attarian et al., 2005), the interneuron inhibition and motor command
(Cantello et al. 1991; Barardelli et al., 1996).
44
In therapeutic, the long lasting effects induced by rTMS can also represent important interests. The
depressing and potentiating effects of rTMS can be used to restore abnormal activity in areas or
networks and lead the system to normal function. The treatment of many neurological and
psychiatric troubles by TMS is extensively studied and evaluated leading to the design of therapeutic
protocols. Such protocols of stimulation complement and potentiate the existing therapies or replace
them when they appear to be ineffective – e.g. in drug resistant patients. The two main syndromes in
which TMS has been used are major depression and auditory hallucinations in schizophrenia (Aleman,
2013). The rTMS strategies developed for the cure of depression are based on a stimulation of the
left DLPFC with high frequency rTMS (George et al., 1997; George et al., 2010). Nonetheless,
inconsistency in the treatment output raises controversy on the efficacy (Speer et al., 2000; Speer et
al., 2009; Speer et al., 2000; Speer et al., 2009; Eche et al., 2012; Speer et al., 2013) and stresses the
need for more investigations in order to understand and design efficient paradigms. Depressive
patients show generally emotion-related symptoms (e.g. Bediou et al., 2005a; Werner and Moulds,
2012; Dillon and Pizzagalli, 2013) with characteristic anhedonia (Rottenberg, 2005) and amplified
response to sad stimuli for instance (Rottenberg et al., 2005). The studies using TMS to dismantle the
emotional processing as we did in our study can bring potential new perspectives and targets in the
thematic of depression treatment. In drug-resistant auditory verbal hallucination, low frequency
rTMS is used in order to inhibit the language comprehension areas and showed some results of
significant improvements (Slotema et al., 2013; Chibbaro et al., 2005; Hoffman et al., 2005; Poulet et
al., 2005; Brunelin et al., 2006). This observation is though mooted as comparable studies also found
poor or no changes (e.g. Slotema et al., 2011) and the improvements did not last over one month
after the rTMS treatment. Emotional processing is also disturbed in schizophrenia (Edwards et al.,
2002; Bediou et al., 2005a; Bediou et al., 2005b) and its investigation, especially using TMS, can also
provide interesting insight for new treatments using brain stimulation. Out of these two main
diseases, therapeutic perspectives for TMS are studied and developed in other pathologic domains.
Many studies investigated the potential of rTMS in pathologies related to pain perception and
chronic neuropathic pain (Khedr et al., 2005; Lefaucheur, 2006; Saitoh and Yoshimine, 2007;
Lefaucheur et al., 2006; Reid et al., 2001; Sampson et al., 2006; Avery et al., 2007; Borckardt et al.,
2007; Short et al., 2009; Martin et al., 2013) as well as fibromyalgia (Sampson et al., 2006; Short et al.,
2009; Li et al., 2013; Mhalla et al., 2011) or even rheumatology (Pérocheau et al., 2013). The
treatment of migraine by magnetic stimulation has also encountered an important development
(Misra et al., 2013; Lipton et al., 2010; Lipton and Pearlman, 2010). On a different field, TMS is
experimented as a method for curing the craving behaviour (Li et al., 2013a; Li et al., 2013b).
Interestingly, the therapeutic use of TMS touches the question of durability of the effects while the
TMS treatment is off.
45
Finally recent use of TMS in preoperative functional mapping represents a promising clinical
application. In replacement or in complement to direct current stimulation (DCS), navigated TMS
could be used to define the cerebral areas involved in functions of particular importance – i.e.
motricity and language – in case of patients with tumour or drug resistant epilepsy scheduled for
surgery. Hence, trains of TMS pulses are used in order to create virtual lesion and reproduce
comparable effects of the DCS on motor (Pitch et al., 2011; Weiss et al., 2012; Coburger et al., 2013)
and language areas (Lioumis et al., 2012; Tarapore et al., 2013; Rösler et al., 2013). Obviously, the
localisation of the target and the assistance in the positioning of the coil are of major importance
with independent or integrated neuronavigation systems. Interestingly, the regions tested with
navigated TMS are not restricted as it is the case with the DCS. Though limited in depth, the TMS
could be used to test any place of the cortical surface and potentially – e.g. in bilateral assessment of
language (Rösler et al., 2013). More exhaustive investigation of the areas could be envisaged
compared to a DCS procedure and analysis of the responses of the patients could be done offline
more precisely.
5. Controls
The control issue is a crucial point when drawing conclusions from TMS experimentation.
Due to the multiple aspects of TMS comprising the expected focused neurophysiological effects but
also the indirect peripheral effects, the perfect control for TMS is difficult to design. It is although
important in TMS experimentation to control the placebo effect of TMS, or more generally the
effects related to the non-neural aspects of stimulation. Placebo response needs to be evaluated
(two recent examples Bae et al., 2011; Jelić et al., 2013) while non-neural effects – i.e. peripheral
sensation and sound, could have potentially important influence on results in event-related
paradigms. Various options have been proposed to answer this issue.
5.1 Control TMS
There are three main manners to control the TMS effect: stimulate other sites, sham design
stimulation and sham coils, or control TMS parameters.
The stimulation of another site, a control site, is intended to reproduce comparable peripheral nonneural effects of TMS and induce a similar magnetic charge to the head without stimulating the
target of interest. The results – behavioural, therapeutic or imaging changes – of this stimulation are
then compared to the results from the stimulation of the target of interest. It may be named active
control site, as the area chosen to be the control receives the same active protocol of stimulation as
46
the tested area. However, even if a priori it does not produce the direct neural effect on the tested
target of interest, it will induce a certain effect on the control site which is not always neutral and
could lead to interference with the conditions studied in the experiment. Moreover, as it is the case
for the site of interest, the stimulation of any site could propagate to its connected areas, being in
non repetitive stimulation (Ilmoniemi et al. 1997; Massimini et al. 2005; Parks et al. 2011; Casali et al.
2013; Ragazzoni et al. 2013) or repetitive TMS (Valero-Cabre et al., 2005; Thiel et al., 2006). The
vertex position on the top of the skull is a common site for active TMS control. The beam of maximal
magnetic field is considered to fall between the two hemispheres in the motor and somato-sensorial
representation for lower limbs. This region is thought to be irrelevant in many cognitive studies and
therefore constitutes often a good candidate as a control site. This solution has been favoured in the
case of our two studies presented in this report. The cortex under the vertex is thought to have no
link in the considered tasks.
Sham stimulation as a control for TMS is done by applying a non effective stimulation on the targetarea of interest. This could be achieved using an active TMS coil in contact with the head of the
subject but tilted with specific inclinations to have a minimal magnetic impact to the brain.
Noteworthy the actual cortical effect of this procedure is not null and has been shown to be
significantly higher with other orientations according to a study using concomitant intracortical
recording in rhesus monkeys (Lisanby et al.., 2001) or EEG in humans (Bonato et al., 2006).
Consequently, some inclinations – e.g. orthogonally to the head surface – are favoured against others.
The other possibility to apply ineffective stimulation is to use a specific sham coil that is designed to
be ineffective. In this case the coil is shielded from a magnetic point of view and no magnetic field is
induced when the electric current flows through it. This solution is thought to eliminate any magnetic
effect but still reproducing the normal position of the coil in contact with the head, its appearance,
its sound and possibly some of the somatosensory sensations due to the electric induced movements
of the wires in the coil. In both cases, the scalp sensation is different for the subject because of the
lack of the peripheral stimulation at the skin surface and in the muscles below the coil. Concerning
this missing aspect, solutions are developed to reproduce those effects more accurately with
additional low electric current stimulations (Rossi et al., 2007, Arana et al., 2008).
A third manner of control in TMS experimentation consists in changing the parameters of stimulation.
These changes could refer to the timing for the pulse launch, the frequency of stimulation, the
intensity (to study a correlation of the effect size with the intensity), the waveform, or the
orientation of the magnetic field. Thus one controls for the specificity of one parameter of the
stimulation in the output effects. This is an excellent control as the same area is targeted by TMS and
peripheral effects are generally comparable. However, the subjective effects could eventually vary, in
47
particular in the case of intensity or frequency changes. In this case the conclusion could misleadingly
reflect effect due to one side effect of TMS.
5.2 Control conditions
TMS effect can be controlled acting on the TMS as developed above, however different
conditions independent of TMS can also be used to control a TMS experiment. Although the goal is
not a direct control of TMS effects but rather a control of the specificity of these observed effects.
Specificity could refer to different conditions or different populations for instance. Does the TMS
induce the same effect according to the conditions? In a behavioural study, one could act on the
tested conditions. This is done with the inclusion of multiple conditions in the main task. It allows
investigating various behavioural effects according to the conditions. In some cases it is necessary to
implement an additional and dedicated task to evaluate the specificity of the TMS in other
behavioural dimensions. For instance in a study using visual stimuli, it would be interesting to use
different visual stimuli with other dimensions – e.g. words vs. image or neutral vs. emotional – to
assess the specificity of the effects for one type of stimuli in comparison to the other stimuli.
Differences in term of the nature of the stimuli could also be envisaged as a control condition – e.g.
auditory vs. visual stimuli or picture vs. movie. In the two studies presented in this report, the control
conditions of the task were carefully chosen and allow for interesting disentangle of the results. In
the study on facial expression the effects were tested in terms of specificity towards emotional
classes (Rochas et al., 2013). In the study on word perception, the inclusion of neutral and emotional
words in the stimuli allowed the effects of TMS to be differentiated between the two categories
(Rochas et al., 2014). Otherwise, in patients for instance when TMS is supposed to have therapeutic
long-term effects, the panel of different symptoms together with other evaluation measures could
also assess the specificity of a TMS treatment.
6. Localisation and neuronavigation
Whatever basic hypothesis is tested, the target-area is one of the crucial points to be
properly defined in a TMS experiment. The expected effect which is related to the function hosted in
the target-area constitutes the main part of the experimental hypothesis. For an accurate
experiment, the target has to be delineated as precise as possible, and then the TMS coil has to be
placed precisely and consistently over this specific target-area. The choice and definition of a target
could be based on assumption based on existing literature, and/or on individual or group imaging
data. Pros and cons are dependent on the various options chosen (Sparing et al., 2010).
48
6.1. Target definition based on existing literature
The choice of the TMS target-area could be based on assumptions derived from previous
imaging studies (fMRI, PET, EEG, MEG, NIRS...) in the literature on the topic of interest. While fMRI
may provide precise locations of the area of interest, EEG or MEG for instance may give in addition
relevant time information that is particularly useful in the case of event-related TMS (e.g. in Rochas
et al., 2013b). TMS location could also be based on results of neuropsychology studies on patients
with brain lesions. In this case the TMS may reproduce transiently in healthy participants the
observed effect of the lesion. The a priori information could also come from intracortical
electrophysiology where precise information about both location and time windows of interest of
the relevant activity is available (e.g. in Rochas et al., 2013a based on Krolak-Salmon et al., 2006). It is
obvious that a target based on literature from group studies of other subjects reported in
standardized brain models will be blurred and imprecise which increases the risk that the actual
magnetic stimulation, though in the most likely activated region in the group, falls aside the
particular point susceptible to produce the expected effects in the particular individual under study.
The choice of a TMS target is constrained by and dependent on the contrasts of the different studies
available in the existing literature. When no appropriate target choice could be deducted from
literature, the TMS experimenters should perform a dedicated localiser phase.
6.2. Target definition from individual imaging
There are two levels of images for targeting the area of interest in the individual brain, an
anatomic level and a functional level. In these two situations a localiser is performed for each
participant – anatomically or functionally – and aims at a personalised localisation of the area of
interest. In this way the definition of a target-area, even when chosen based on the literature, can be
anatomically established in the individual subject. The part of the brain – e.g. a specific gyrus, a point
defined by standardized coordinates or a Brodmann area – is pointed as a target and is precisely
identified on individual structural MRI. This is the method used in the two works presented in this
present report. Even more sophisticated is the precise individual localisation of the target in a
dedicated functional localising phase using either an fMRI or a EEG/MEG exam (in EEG/MEG source
localisation is required). Generally speaking, the functional localiser has to light up the areas of
interest using appropriate contrasts; the relevant areas are subsequently defined as individual region
49
of interest in the MRI of each participant/patient. Note that connectivity information could enable to
reach deep connected structures or to expect some effects coming from the connected areas.
The use of an individual anatomic localiser resolves the problem of inter-subject anatomic variability
and adjusts the stimulation on the correct structure of the cortex but it assumes a consistency in
relevent activations across individuals. The individual identification of functional target allows a
better control of variability between subjects not only in terms of anatomy but also in terms of
function (Sack et al., 2009).
6.3. Positioning methods
Once the target is defined and its position identified in a specific and exploitable space – a
standardised space or the individual space – the coil has to be placed on the head as percise as
possible. Taking into account the fact that the target of TMS is on the cortical brain surface and that
the scalp, skull and meninges are not transparent, the placement of the coil to reach the desired
point with the induced magnetic field is not so obvious.
Anthropomorphic placement based on diverse scalp measurement systems was applied for areas
with an available well documented anatomy – e.g. the primary motor cortex – and in studies where
the precision did not matter – e.g. with a large or a circular magnetic coil. Even if scalp
measurements demonstrated acceptable accuracy for well known areas compared to frameless
neuronavigation systems (e.g. Broca area in Weiduschat et al., 2009), the procedure has also proven
evident bias in the placement of the coil over different targets such as DLPFC (Nauczyciel et al., 2010;
Ahdab et al., 2010).
Nowadays, neuronavigation systems based on image reconstruction of the individual brain are used
for this purpose. Neuronavigation systems combine the actual head and coil with their virtual models.
Using a pointer carrying its own landmarks, the neuronavigation procedure consists first in a
coregistration of the landmarks placed on the participant head in the real space with some marks on
the subject MRI. The same procedure is done for the coil itself with its 3D model. The localisation of
the landmarks is done by a spatial triangulation. The two main systems of landmark tracking utilise
ultrasounds or infrared light. With ultrasounds, little speakers work as emitter landmarks and
microphones as receptors. With infrared light, cameras act as an emitter and as a receptor, and the
reflectors represent the landmarks. In comparative studies, the use of functional localiser with
neuronavigated TMS seems to give the best results among the different possible couples of
“definition of location” of the target and “placement method” of the coil (Sparing et al., 2008; Sack et
al., 2009). Neuronavigation systems also allow viewing the reconstructed image of the brain macro
50
structures – i.e. gyri and sulci. This could help to place and maintain the coil in an appropriate
position according to the orientation of the targeted structure (Cincotta et al., 2010), and also to
avoid neighboured structures.
7. Safety
TMS is a non invasive technique used for decades in research and clinical protocols. By the use of
high voltage currents, the TMS has to be carefully operated for both the participants and the
operators. Due to its particularity of being a magnetic field generator, the use of TMS implies also
supplementary safety aspects to avoid any hazardous interactions of the participant or the
environment with the magnetic field. Any inadequate ferromagnetic material or electric device has
to be kept away from the stimulator coil when working. Finally, the nature of the stimulation
conveyed by the generation of micro currents in the brain tissue should be considered carefully and
encourage the experimenter to pay special attention at the participant antecedents and not exceed
the recommended charge of stimulation. The TMS experiments have to be in conformity with the
safety and application guidelines.
Thus specific guidelines and handbooks have been developed, updated and hence lead the clinicians
and researchers to a safe and appropriate use of the TMS technique (Pascual-Leone et al., 1993;
Wasserman, 1998; Rossi et al., 2009; Pascual-Leone et al., 2002). In addition, validated screening
questionnaires are commonly used to evaluate the safety of inclusion of the participants and patients
(Rossi et al., 2011).
7.1. Direct effects due to the TMS magnetic field generation
The flow of high electric currents of thousands of amperes and volts in the TMS machine and
coil is the first preoccupation to be considered. Like any electrical devices, it has to be used with
respect of safety rules of the constructor and in an “electricity-friendly” and safe direct environment
– e.g. no humidity, no proximity with liquids, non containment and fire protection.
Secondly the high magnetic field induced by TMS could raise some issues. Because this magnetic field
is not continuous and limited in space around the stimulator coil – no static magnetic field, the dose
exposure for participants and operators would be rather low for single pulse or even for rTMS
procedures when considered in absolute value – e.g. 1500 biphasic pulses of 300 µs make a total of
0.45 seconds exposure. Nevertheless, the indirect effects of one pulse could last for a longer than its
simple pure duration.
51
Using time-varying magnetic field, the finality of the TMS is the induction of electric micro eddy
currents in the conductive nerve tissue. At the level of the brain tissue the induced electric field is on
the order of a maximum of 150 V/m (Epstein et al., 1990) depending on the magnetic field amplitude
and waveform. Hence the TMS-induced electric currents are considered safe for the nerve tissue.
However the amplitude of the induced voltage is also dependent of the considered medium
impedance – i.e. higher conductance meaning higher voltage current. The relative higher
conductivity of other materials – e.g. in scalp electrodes, implanted epidural or deep brain electrodes,
in the lead wires, in cochlear implants or in any other electrical devices – makes them possibly act as
antenna inducing higher currents and could lead to hazardous consequences for both participant and
material. Non ferromagnetic material and special designs have been developed to counteract both
magnetic and electric interaction with TMS (Ilmoniemi and Kicić, 2010). Concerning implanted
electrodes, ex vivo studies with phantom models demonstrated safe and ineffective current
induction by TMS in the electrodes (Kumar et al., 1999; Kühn et al., 2004; Schrader et al., 2005) but
unintended possible high electric currents generated in lead wire loops (Rossi et al., 2009; Deng et al.,
2010; Shimojima et al., 2010). Furthermore, supra-threshold motor activity were shown in response
to TMS-induced current in patients with electrode leads connected to implanted stimulators in a few
in vivo studies (Kühn et al., 2002; Hidding et al., 2006).
Another consideration about conductive material is that the induced electric currents in return make
this material subject to the Lorentz force law. These forces induce movements of a material
traversed by time varying currents. This can be a particular issue in patients with metallic implants.
However different studies have shown the inocuity of these forces on regular clinical implants often
made of titanium (low conductive and non-ferromagnetic), as surgical metallic skull plates
(Rotenberg et al., 2007; Rotenberg and Pascual-Leone, 2009), clinical aneurism clips (Barker, 1991),
or again implanted electrodes (Shimojima et al., 2010; Golestanirad et al., 2012).
Contrary to electric stimulation, the temperature increase induced by magnetic stimulation in the
living tissue is very limited (Ruohonen and Ilmoniemi, 2002) and curbed by the abundant blood
vascularisation of the brain tissue. The heating of the brain is below the accepted threshold studied
in other magnetic imagery techniques (Brix et al., 2002) but the heating of material in contact with
the skin or implanted in the head could induce indirect burn and injury if not considered carefully.
Specific electrodes - - have been developed to perform EEG combination (Roth et al., 1992; Virtanen
et al., 1999, Ives et al., 2006) and the implanted material used in clinic does not reveal heating
problems (Rotenberg et al., 2007; Rotenberg and Pascual-Leone, 2009; Hsieh et al., 2012).
If not already done, the current, movement and heat induction in any new implanted materials
should be tested in an ex vivo montage reproducing the conditions of stimulation then reported.
52
7.2. Indirect effects of TMS on biological systems
An epileptic seizure is due to an abnormally high synchronicity of a group of neural cells of the brain
and could lead to symptomatic outcomes ranging from loss of awareness and alteration of cognition
to uncontrolled tonic and clonic muscular activity. Indirectly related to the local but massive
activation of neural cells – i.e. by driving – or to the long-term changes in excitability in the time
following the stimulation, the risk of seizure represents the major hazardous event in term of
severity during TMS procedure. The number of TMS-induced seizures – nineteen cases (sixteen
reviewed in Rossi et al., 2009; Kratz et al., 2011; Edwardson et al., 2011; Vernet et al., 2012) – is quite
limited regarding the total amount of TMS procedures. Importantly, the majority of these few cases
of seizures occurred outside the safety guidelines while the others were a posteriori found in
conditions known as pro-epileptic – sleep deprivation or alteration, high cortical excitability
threshold, pro-epileptic drugs. Epileptic patients and patients with diseases related to positive
sensory phenomena (tinnitus, auditory hallucinations, visual hallucinations, pain syndromes, visceral
pain, migraine, or fibromyalgia) constitute population with potential high risk of seizure induction.
However Bae and colleagues (Bae et al., 2007) reported only 1.4% of seizure event in a population of
280 epileptic patients undergoing rTMS treatment. In epileptic populations, particular attention
should be paid when attributing a seizure induction to TMS because of spontaneous high seizure
susceptibility and possible intake of drug that may change the excitability of these patients. Finally in
some cases, proper epileptic seizures were even doubted and possibly mistaken with syncope.
Syncope is a transient loss of consciousness accompanied by a loss of postural tone. The cause of
syncope is a momentary cerebral hypoperfusion due to nervous and bodily adverse events. Contrary
to seizures, syncopes are not triggered by cerebral electrical disturbance but rather by psychological
stress and anxiety induced by the TMS. It is a less severe adverse event compared to seizure however
its spontaneous occurrence is higher in a given population as well as during TMS procedure
(hundreds of syncopes not always reported versus 19 cases of suspected seizures). They are
sometimes hard to differentiate from proper epileptic seizure (Epstein, 2006).
Sound produced during TMS is due to the time varying high magnetic field inducing movement and
shocks in the wires of the loop. This brief sound artefact produced at each pulse lasts only hundred
of microseconds (Tringali et al., 2013). Actually the produced noise could easily reach 120 dB sound
pressure level at the peak for usual intensities of stimulation (Starck et al., 1996; Tringali et al. 2012;
Tringali et al., 2013). However the briefness of the sound lessens the discomfort and conceals its
toxicity (Starck et al., 2003). Moreover the stapedial reflex needs 25 ms to be engaged in case of high
53
intensity auditory stimulus to protect the inner ear. Hence the pulses in TMS with a peak-like
waveform would be more damaging compared to a continuous sound with similar sound level
(Forget, 2011) and could induce hearing loss (Johansson, 1980). Hearing protections are highly
recommended for both participants and experimenters.
The practise of TMS cannot be dissociated from possible adverse effects as peripheral sensations and
even light pain. These symptoms are generated by the stimulation of muscles and nerve terminations
at the surface of head. Nausea has also been reported after TMS sessions. Except for local discomfort
and pain, no direct relation has been established between TMS parameters and mild adverse effects
in a recent review by Maizey and colleagues (Maizey et al., 2013). Interestingly they did find only
marginally significant differences in occurrence of mild adverse effects between real and sham
stimulation and there was no difference in types of the adverse events.
Other biological effects related to safety concerns are currently discussed – e.g. neurotransmitters,
endocrinology, immune system, autonomic function. For instance in a mice study, rTMS modified the
expression of mRNA of monoamine transporter (Ikeda et al., 2005). In humans, recurrent sessions of
rTMS for patient treatment would cause significant changes in anatomical structures (Peng et al.,
2012) and microstructures (Baeken et al., 2012). These effects are rare and related in particular
procedure of TMS however more studies are needed in this domain.
IV. THE USE OF ELECTRICAL NEURONAL SIGNALS TO GUIDE TMS
As already mentioned in several parts of the present report, electromagnetic imaging, and
more particularly the EEG, represents an interesting modality to be combined with TMS. The
recording of electromagnetic signals produced by the brain is opposite to the TMS where
electromagnetic current is brought into the brain. Because of the high temporal resolution EEG
allows directly the study of the induced rapid changes of the electromagnetic stimulation. The
temporal resolution of the EEG also allows the definition of the moment in time where certain
functions take place and thus allows the determination of the time point of TMS stimulation after a
stimulus. In order to also allow defining the location of stimulation with EEG, electrical neuroimaging
based on source localisation of high-density EEG is required. Alternatively, and as shown in the first
study of this thesis, intracranial recordings with depth electrodes in epileptic patients can be used to
guide target localisation for TMS. This section describes these methods and its implications for TMS
studies.
1. Intracranial electrodes for target localisation
54
Intracranial electrodes are electrodes directly implanted in the human brain. As this includes
invasive procedures it is done in particular clinical settings. The recording from the implanted
electrodes is called electrocorticogram (ECoG) or intracranial EEG (iEEG) or depth electrode
recordings. The procedure is called intracortical or deep brain stimulation (DBS) when the purpose is
to convey electric current in the implanted brain structures in order to activate or disturb them.
Recording and stimulation are often combined in the same implanted electrodes. This was the case
in the depth electrode study of Krolak-Salmon and colleagues (Krolak-Salmon et al., 2006) that
inspired our first TMS study presented hereafter (Rochas et al., 2013).
1.1. Recording
Depending on the electrode type and placement, intracranial electrodes capture the electric
activity of more or less large number of cells. Thus the signals rather represent local field potentials
than single unit and spike activity. Local field potentials could be analysed in terms of frequency band
power or in terms of evoked potentials. While frequency analysis can be performed on spontaneous
recordings as well as in stimulation situation, evoked potential analysis requires repeated
presentation of stimuli in whatever sensory modality.
1.2. Deep brain stimulation
Deep brain stimulation (DBS) consists in the injection of electric currents in a limited zone of
the brain directly through the intracranial electrodes. It generally produces a transient disruption of
the function in which this area is involved. The effect of TMS could be similar being it online in an
event related paradigm or offline with rTMS protocols. DBS can also induce activity and overt
responses (particularly motor manifestations) induced by stimulation of a brain area. Such effects
can also be induced by TMS. DBS or more general electrocortical stimulation is used in Parkinson’s
disease as well as in presurgical epilepsy evaluation, where electrocortical stimulation serves to map
the individual brain functions.
1.3. Target localisation.
An important consideration when using intracranial data to guide TMS target localisation in
healthy subjects is the fact that the results come from patient populations, meaning generally a
limited population and possible altered structures and functions. The specific location found in one
or several patients has to be transposed and defined in a brain of the general population to be
55
practicable with TMS. In case of a patient brain without lesion, the area of interest can be expected
to be similar in healthy participants. However in the case of patients with brain lesions,
reorganisation of brain structures and functions may have taken place and the area to be stimulated
in the healthy participants would not be necessarily at the same place. This is particularly true, when
the area of interest identified in a patient is close to a diseased area or in a network suspicious to be
functioning abnormally. However TMS in healthy participants could be a way to check the validity of
the intracranial results by the induction of a virtual and transient lesion of the area investigated. In
fact TMS is the perfect tool to reproduce the results obtained in DBS. Note that TMS is increasingly
considered as a presurgical functional mapping procedure (see section I.4.3. on preoperative
mapping). Importantly, due to the decrease of the magnetic field with distance, the TMS is applicable
in cortical areas only.
2. EEG for target localisation
Electroencephalogram is a neuroimaging technique based on the recording of the electric
potential on the scalp surface produced by neuronal activity in the brain. The time resolution of this
technique is extremely high in the order of the milliseconds. Generally offline, different levels of
analysis are possible from the analysis of waveforms at single electrodes up to the reconstruction of
the three dimensional distribution of the sources in the brain (Michel and Murray, 2012). As already
mentioned, the combination of TMS with online EEG presents numerous advantages. An offline EEG
experiment provided the space and time information leading to the choice of the target and TMS
protocol in second TMS study presented hereafter (Rochas et al., 2014).
2.1. Evoked potential and waveform analysis
Multichannel EEG recordings consist of multiple waveforms recorded at the different
electrodes as potential difference between the electrode and a reference electrode. These different
waveforms can be analysed for each electrode separately. Hence investigations are performed
electrode by electrode and do not include an integrative view on the spatial pattern of the potential
distribution.
In evoked potential studies, the signals are usually averaged across several similar events. This
reduces the noise and amplifies the signals that are only in the order of a few microvolts in contrast
to the ongoing activity which is around 100 microvolts. Waveform analysis of evoked potentials
typically consists in determination of amplitude and latency of certain evoked potential peaks. These
deflections in the signal of a given electrode are supposed to index specific information processing
56
steps. Importantly, the signal obtained at a single electrode dependents on the signal recorded at the
reference electrode.
The waveform analysis is restricted to information from isolated potentials recorded on specific
electrodes on the scalp and is interesting for rapid analysis of well-known components – e.g. visual
P100, auditory N100, N400, etc. – but fails in the construction of an integrative view of the brain
state by taking advantage of the entire available signals from the all set of electrodes. Thus, for the
use of EEG to guide TMS location, single channel waveform analysis is not very helpful, as the
electrode under which a certain component is maximal does not directly indicate the location of the
generators of this component.
2.2. Topographic analysis
Topographic analysis of EEG recording refers to the analysis of the scalp distribution of the
potential field. Scalp potential maps can be constructed at every time point, representing the sum of
all momentary active neurons in the brain at a given time point. It is important to note that the
configuration of the scalp potential field is independent of the chosen reference electrode in contrast
to the waveforms described above (Geselowitz, 1998; Lehmann and Skrandies, 1980; Pascual-Marqui
and Lehmann, 1993). The pattern of a map is the signature of a brain state at a moment in time. Two
different maps – different distributions – have been generated by different sources in the brain and
might thus represent different brain states; however the inverse does not hold (Vaughan, 1982;
Michel et al., 2004). Hence differences of scalp potential distributions are informative as they
represent differences in brain processing, and analysis methods based on map topographies aim to
identify these different processing stages. Topographic analysis can be applied to evoked potentials
as well as to spontaneous activity (microstate analysis; Michel et al., 2012; Lehmann and Michel,
2011).
Different methods have been developed to analysis scalp potential maps and their differences across
time and between conditions. In evoked potential analysis for instance, topographic analysis of
variance (T-ANOVA) aims at comparing the electric current topographies as dependent variables
using randomisation statistics (Murray et al., 2008; Koenig et al., 2013). Multivariate T-ANOVAs have
also been developed (Koenig et al. 2011). Using permutation tests, the T-ANOVA compares the real
distribution of the map dissimilarity between conditions with the dissimilarity of the distribution of
the random maps across subjects. It allows identifying time windows of topographic divergences
between various conditions and potentially relevant for the studied hypothesis as it was done in the
second study presented in this report (Rochas et al., 2013).
57
Other programs and strategies are available to perform topography-based analysis (Murray et al.,
2008; Brunet et al., 2011). One of these approaches is the microstate analysis, which aims at
identifying time periods of stable topographies over time and differences of these topographies
between conditions. In evoked potentials maps typically show stable configurations for tens of
milliseconds called microstates. Short transitions separate the periods of stability. Stable
topographies typically represent the well-known different evoked potential components. Clustering
methods identify relevant microstates and fitting procedures quantify the presence of these
microstates in each subject. Microstate analysis can be useful for the identification of time windows
where TMS could be applied to eventually interfere with a particular processing step.
Hence the different kinds of topographic analysis identify time periods of specific brain state
potentially relevant for the application of TMS.
2.3. Source localisation
Electrical source imaging (ESI) uses inverse solution methods in order to calculate and localise
the potential sources of brain activity for any moments from multichannel scalp EEG (Michel et al.,
2004). Directly derived from topography maps, the solution of the inverse problem takes into
account all the electrodes and is also independent of the reference. Importantly it keeps the
extremely high temporal resolution of the EEG for source localisation and allows the chronologic
modelling of brain activity. The 3-D reconstruction of the sources or generators of activity in the
brain can be done on an averaged head model – e.g. MNI brain – from which standard coordinates
are extractable or directly from the MRI data of the individual subjects.
ESI is a mandatory step in order to use EEG recording for defining the target-area for TMS. It provides
location of sources in time and constitutes a particularly well adapted method of localiser for eventrelated TMS paradigms when it is based on ERP. Thus the topographic analysis described above
together with electrical source imaging provides a unique possibility to identify the time period as
well as the brain area to apply single pulse TMS to interfere with a certain function. On the other
hand, TMS allows for the validation of the results of topographic evoked potential analysis by
eventually demonstrating causal relationship between the time and location of the stimulation and
the studied function. This approach was used in the second study of this thesis.
58
V. HYPOTHESIS
The work of this thesis raises both methodological and experimental issues. Concerning the
methodological side, the work assessed if electrophysiological recordings can be used to guide eventrelated TMS for the modification of the perception of emotions. We used either intracranial
recordings or high-density scalp EEG with source imaging to define the location and the timing of
stimulation and hypothesise that such advanced methods will very precisely guide the subsequent
TMS studies. Besides the methodological aspects of the thesis, we were interested in the perception
of emotional faces and words that are major vectors for emotional communication. We asked in our
studies when in the processing stream the emotional information of these stimuli is decoded. We
were interested in brain areas outside the classical areas associated with emotion processing. We
wanted to know if these cortical areas are relevant in emotional processing and with what exact role.
More precisely, what is the role of the premotor cortex in the processing of emotional facial
expressions? Does it show mimicry-like responses? What is the role of the right hemisphere in
emotional word processing? Does it work alone or in cooperation with the left hemisphere?
In short, the present work assesses the following hypotheses:
- The combination of electrophysiology prior to event-related TMS allows to precisely localising
emotion processing in time and space and in various modalities (study 1 and 2). This combination
provides a unique access to brief and/or early activations implicated in emotion perception.
- The stimulation of the premotor areas leads to alteration of specific processing of emotional facial
expression related to the mirror neuron system (study 1). The mimicry system is necessary and has a
real functional role in the correct perception of emotions. Its implication should be dependent on the
category and intensity of emotional stimuli.
- The stimulation of the right hemisphere at early time points leads to alterations of emotional word
perception (study 2). The right hemisphere plays a role in the early processing of emotional written
words, in cooperation with the left hemisphere.
VI. DESCRIPTION OF THE ARTICLES
1. Disturbance of facial expression recognition by TMS
1.1. Summary of the results
This study (Rochas et al., 2013) used a task of facial expression recognition with pictures of
faces in different intensities of happy, angry and fear expression. The results of the event-related
TMS confirmed the implication of the left pre-supplementary motor area in the recognition of facial
59
expression. Compared to the stimulation of the vertex as a control condition, the stimulation of the
left pre-supplementary motor area by 10 Hz train of pulses released at the offset of the presentation
of the pictures led to a specific disruption of the recognition of happiness expression. The accuracy
was decreased for the recognition of happiness while no effects were observed for the recognition of
anger or fear.
1.2. Contributions to this work
In this project, I participated in:
- Pilot study in order to test the stimuli material and the behavioural task on independent volunteers
- Design and programming of the task and setup of the paradigm of event-related TMS together with
the neuronavigation system
- Performing of the experiments for half of the participants of the study
- Analysis of behavioural measures
- Writing of the manuscript
- Submission process
2. Lexical detection of emotional words
2.1. Summary of the results
This study (Rochas et al., 2014) gathers an event-related EEG experiment and an eventrelated TMS experiment on two independent groups of participants with the same task of lexical
decision with simultaneous bilateral presentation of letter strings – neutral or emotional words
paired with pseudo-words. The EEG results revealed two periods of interest in the early processing of
lexical decision. These periods of interest showed specific brain states according to the side of
presentation and the emotionality of the detected words. Specific activations were found notably in
the right temporoparietal junction for emotional words especially when they were presented in the
left visual field. When compared to the vertex stimulation, the stimulation of the left or right
temporoparietal junction led to an interaction with a reaction time increase for emotional word
detected in the left visual field.
2.2. Contributions to this work
60
In this project, I participated in:
- Pilot study in order to test and adapt the stimuli material and the task on independent volunteers
- Design and programming of the behavioural tasks
- Setup of the paradigm of event-related TMS together with the preparation of the neuronavigation
system
- Performing of the experiments for all the participants of the EEG and the TMS experiments
- Analysis of event-related potentials from the EEG experiment
- Analysis of behavioural data from the TMS experiment
- Writing of the manuscript
- Submission process
3. Contribution to additional works
I also participated in a study of investigation of facial mimicry processes in emotion recognition using
TMS and EMG. I notably gave input for the setup of the paradigm of theta burst stimulation and of
the experimental procedure. I managed the setup of the neuronavigation system and effectively
participated in the practice of TMS.
Korb S, Malsert J, Rochas V, Rihs TA, Schwab S, Rieger S, Grandjean D, Niedenthal P. The role of the
motor and somatosensory cortices in facial mimicry. Analysis in progress.
I participated in a study on attention to emotions combining repetitive TMS and fMRI. I notably gave
input for the paradigm of theta burst stimulation and for the experiment procedure and control. I
also managed the setup of the neuronavigation system and effectively participated in the practice of
TMS.
Malsert J, Rochas V, Rihs TA, Rieger S, Pichon S, Vuillemier P. Probing the role of the FEF in emotional
attention with TMS and fMRI. Analysis in progress.
I participated in the beginning of a longitudinal EEG study on patients with 22q11 deletion syndrome.
I notably helped for the setup of the EEG protocol and the procedure of eye tracking.
Rihs TA, Tomescu MI, Britz J, Rochas V, Custo A, Schneider M, Debbané M, Eliez S, Michel CM. 2013.
Altered auditory processing in frontal and left temporal cortex in 22q11.2 deletion syndrome: a
group at high genetic risk for schizophrenia. Psychiatry Res, 212(2):141-9.
61
VII. DISCUSSION
1. Foreword
The brain is a prior actor in emotion and can be considered as “the” emotional organ. One
can note that it is though in permanent interaction with signals from the body. Integrated models of
emotion processing establish a relation between brain and body. For instance, body language is part
of the emotional perception while hormones, secreted by both brain and body, influence the
emotional state. The brain expresses its emotions by the body and processes them through the body.
Thus one could have chosen to act on the body to alter emotion processing, but we were interested
in brain modifications using TMS. We investigated the modification of emotion perception acting on
areas of the cerebral cortex.
Facial expressions as well as words come up to mind immediately when talking about emotions.
Faces and words are two major vectors of the communication of emotions. Considering the ubiquity
of emotions across numerous structures in the brain, the disruption of their perception by TMS can
appear to be trivial as they could implicate any brain regions. The disruption of the normal function
of any cortical area by neurostimulation might modulate the perception of emotions through these
modalities. However, if almost any area could potentially be involved in parts of processing linked to
emotions, these processing are specialised in the different aspects and dimensions of emotions and
show more or less significance. The perturbation of fine compounds of their perception seems thus
trickier. Also the many parallel systems involved in their processing could make them robust and
hard to disrupt, especially when acting on additional and non-usual processors of emotions; what
could be called atypical emotional areas. The perturbation of the chosen areas in the present studies
led nevertheless to the modification of emotion processing in general, and more precisely in one
compound of the perception of emotions at a certain timing. The targets were not parts of the limbic
system or some well-known areas devoted to emotions such as the insula. A part of the premotor
system was chosen in the first work presented here (Rochas et al., 2013) and an associative area of
the right hemisphere in the second one (Rochas et al., 2014). As we will discuss, the premotor area
was tested for the decoding and quantification of emotional movements; the right hemisphere and
especially the temporoparietal junction (TPJ) was foreboded in the really precocious emotional
categorisation of words. These two systems were considered as highly specialised processors of
emotionality and their disruption in relatively early time successfully modified the behavioural
response of participants towards emotional stimuli.
The modification of distinct and precise aspects of the processing of emotions was only possible
thanks to prior investigation of the emotional functions with electrophysiology. The crucial choice for
62
where and when to stimulate was indeed directed by these electrophysiological studies. In the first
presented work (Rochas et al., 2013), the results from an intracranial EEG recording in a drugresistant epileptic patient (Krolak-Salmon et al., 2006) led to the formulation of the main hypothesis:
the left pre-SMA is implicated in the processing of facial happiness. In the second presented work
(Rochas et al., 2014), the preliminary EEG study in independent healthy volunteers provided
information to track early steps of emotional processing of written words in the right hemisphere. It
brought basics to hypothesise the role of the right TPJ in the early processing of emotional words and
led more precisely to the targeting of the angular gyrus (AG). Hence, one can conclude that even if
the brain is emotional, one can hardly follow a try-and-see procedure by testing any area in any
emotional task with TMS. Here, a priori is a prerequisite. We will discuss the contribution of
preceding electrophysiological studies to guide event-related TMS.
2. Behavioural measure and event-related TMS paradigm
Considering the modification of the perception of emotions, several questions could have
been envisaged and consequently several strategies could have been chosen by using TMS. For
instance, we could have investigated the effects of the disruption of different areas studying one
aspect of perception of emotions. Still with one aspect of emotion perception, we also could have
looked at different effects according to the diverse possible TMS paradigms on one unique cortical
target. Instead the presented works used a similar stimulation paradigm – i.e. event-related trains of
biphasic pulses – on different cortical targets in different behavioural tasks. While both of them
explore the visual domain, the two functions targeted in these works were tested in different
modalities of emotion perception that are the facial expression and the written words. Also we could
have looked at the effects of event-related TMS on brain activity with a combined neuroimaging
monitoring. However we were interested in the causal implication of the tested areas in the
behavioural compound of emotion perception. We wanted to see if the initial modification of the
targets would be responsible for relevant alterations of the behaviour. Consequently we used
behavioural tasks to assess these functions on different emotional dimensions according to the
expected roles of the two targets.
In both studies, we were interested to show an implication in early steps of processing. The use of
event-related TMS allowed assessing the involvement of target-areas at early time points. The
disturbance in the early period after the presentation of the emotional stimuli showed primordial
mechanisms of perception – differently from the preparation before the perception or later stages of
processing. These effects could not have been studied with rTMS for instance, due to long-term
63
effects on the target-area. In this sense, the event-related TMS is more informative about the nature
of the mechanisms that are involved. In addition the event-related paradigm of stimulation might act
on precise targets and avoid possible diffusion of disturbance unlike rTMS. The long-term effects of
rTMS on a target-area are possibly responsible for long-term changes of large connected networks
(Thiel et al., 2006). This mechanism of plastic adaptation is unlikely to occur in short delay of the
induced event-related TMS disturbance. Hence its action is more restricted in time and also in space.
However the non-diffusion of single pulse TMS is not always true. In fact, some studies have found
activations induced by TMS in remote areas that are connected to the originally targeted site
(Massimini et al., 2005; Casali et al., 2013; Ragazzoni et al., 2013). Effects of event-related TMS on
remote connected areas might be limited thanks to the different threshold gates constituted by
synapses – of course always depending on the populations of cells stimulated in the original targetarea. The fact that the intensity of stimulation in our two studies were moderate and did not exceed
the motor threshold might also limit the propagation of the stimulation effects through the cerebral
networks. Near-infrared light imaging technique (called EROS) combined with TMS showed
unidirectional propagation of TMS effect from the left primary motor cortex to the right homologous
area from 40 to 48 ms post-pulse but not from the right to the left possibly due to insufficient power
of stimulation over the right M1 (Parks et al., 2012). The addition of a neuroimaging technique during
the TMS performance would have provided interesting information on the potential of propagation
of our stimulations. Unfortunately expertise in this technique was not available at the time when the
studies were performed.
Nevertheless, behavioural measures provide a concrete measure of relevant effects of TMS on the
perception and processing of emotions. In fact behavioural measures give access to the implication
of the stimulated cortical areas in the studied functions. It is a verification of the causal role of the
activation that was observed in the electrophysiological recordings. In the first study, TMS did not
replicate the productive effect of the DBS on the left pre-SMA; neither uncontrolled happiness
reaction was observed nor even a modification of mood. The effects of the TMS were certainly too
moderate to do so but they were nonetheless transcribed in behaviour. The observed alteration in
facial expression recognition in the participants (Fig. 2 in Rochas et al., 2013) was in agreement with
the previous specific evoked potential for happy face presentation found in the patient of KrolakSalmon and colleagues (Fig. 2 in Krolak-Salmon et al., 2006). The behavioural modification in healthy
participants thus confirmed the specific phenomenon previously found in a patient and moreover it
suggested its important role in decoding of the emotional expressions. In the second study, the
behavioural effect was represented by a slowing of emotional word recognition especially when they
were presented in the LVF. This effect was encountered for both the stimulation of the right and the
64
left AG. The behavioural effects again were in line with the results observed in the EEG in the same
task, and more particularly the ESI. In fact a specific activation was found in the right TPJ for the
detection of emotional words compared to neutral words, and for word in the LVF compared to the
RVF (Fig. 4 in Rochas et al., 2014). However these specific activations were constantly accompanied
by unconditional activations in the homologous left hemispheric regions. The TMS effects due to the
disturbance of the left or the right AG may suggest an early cooperation between the two
hemispheres in the processing of emotional words. Alternatively the effect argues at least for a
transfer of the processed information from the right to the left hemisphere. As discussed in the
article the impossibility to isolate the effects from TMS to the left or to the right AG in our results
could also be due to a remote alteration of the functioning of one area by the stimulation of the
connected homologous area – i.e. the right by the left TPJ region. Hence it would not be possible
with this study alone to fully dismantle effects of the two sites of stimulation as both could be
induced by the effective stimulation of one area alone – i.e. the right AG. Again, the combination of
the TMS with online neuroimaging could have answered this important concern. Nevertheless, the
results on emotional word detection after TMS together with the results from ESI point obviously to
a crucial early direct involvement of the right TPJ in the processing of emotionality in written words
especially when they were presented in the LVF. In fact, the interaction in the behavioural results
from TMS in this second study reflects a conditional involvement of the bilateral communication
between the TPJ in the treatment of word material conditioned by their emotionality and LVF
presentation; and a lesser involvement when being neutral and presented in the RVF. This last point
explains the variable implication of the two sides of TPJs observed in EEG raw ESI (Figure 8).
Figure 8: unreported data from the second study of this thesis report, Rochas et al., 2014: Representation of the brain
activations for the detection of neutral or emotional words in the LVF or the RVF at 120 ms after stimuli onset on a
transversal section of an MNI brain model. The colours represent the group average of the normalised z-scores for
electrical source activities between all solution points; with warm colours for the highest spots of activity and cold
colours for the lowest spots of activity. Note that at this critical time point, in the end of the first period of interest
defined in the article, the right TPJ is particularly active when detecting words in the LVF and emotional words.
65
While the paradigms of TMS were similar in the two experiments of this thesis – i.e. early stimulation
with rapid train of pulses, the observed effects of the TMS were of course different and dependent
on the targeted areas and of the considered function in emotion perception. The disruption of the
left pre-SMA induced a specific impairment of the performance of the participants in the recognition
of happy faces without alteration of their reaction times. The TMS has actually influenced a part of
the expression recognition process and made it impossible to correctly identify emotions in some
cases – with possible emphasis for moderate intensities of emotion (Figure 9). The implication of the
premotor system will be discussed in more details below, but it is worthwhile to note the
differentiation found in the behavioural results for a particular category – and possible intensity
range – of emotional expressions. It is not sure if such particularity could have been found with a
rTMS paradigm instead of event-related TMS or in data from neuroimaging studies instead of
behavioural measurements. Concerning the stimulation of the right AG and its left homologous area,
the effects were present in the reaction time but not in the accuracy of detection. This effect on
emotional detection after the disruption of one side of a bilateral structure – i.e. both TPJs – showed
a subtle functional interaction between the two hemispheres. It will be discussed in more details
below. The slowing without decrease of performance during TMS indicates also a probable
processing of emotionality in other networks and certainly in the left hemisphere – in a left
specialised language network excluding the left AG itself.
The combination of event-related TMS and behavioural measure revealed various aspects of two
different systems of treatment of emotional stimuli. It first confirmed the implication of certain areas
which included activations in the electrophysiological recordings, and it even refined the condition of
involvement and of functioning of these areas.
3. Potential of electrophysiology for TMS target definition
Our studies demonstrated that electrophysiological measures can be an accurate indicator of
brain areas involved in emotional processing in a certain task and a guide for assessing effects on
emotional perception with event-related TMS. First of all, event-related potentials provide
information on the location of the stimulation. Indeed the suspected specific activity found in
intracranial recording in the left pre-SMA led to the subsequent specific effect on behaviour in
healthy participants when applying TMS to this area (Rochas et al., 2013). Notably, the activity
extracted from one single electrode contact in a patient was able to accurately guide the TMS study.
In the second study the right hemisphere structure that was part of a bilateral network identified in
EEG source imaging, gave interesting interaction in behaviour when stimulated with TMS (Rochas et
66
al., 2014). Even if the electrical brain activity were recorded on a different population, TMS on the
positions extracted from the EEG group analysis revealed significant effects in emotional perception
in another population. A more direct approach would have been to use the same population in the
EEG and the TMS parts of the study, and to localise the targets for each participant in order to have
an individual adapted positioning of the magnetic stimulation. Unfortunately the majority of
participants were unavailable for the TMS part. Also to avoid a possible learning effect of the task
between the two parts for those who had already done the EEG part, a completely independent
sample of participants was finally chosen for the TMS study. Electrophysiology is of course not the
only manner to find the areas of relevant activity in the brain. EEG is particularly adapted to find
cortical sources, even if nowadays improvement in source determination allows also identification of
subcortical brain structures. However, as TMS stimulation also has limited access to deep brain
structures, the bias of EEG to cortical areas is not a limitation. According to our investigations, the
two electrophysiological techniques employed for target definition were well appropriate and were
certainly comparable to what could be expected with a localisation using fMRI investigations for
instance (Sack et al., 2009).
Importantly, the time is crucial for event-related TMS, reason why event-related potentials for TMS
guidance bring a real advantage over other neuroimaging methods like fMRI or PET with very limited
temporal resolution. Event-related potential analysis measures the time course of brain activity after
an event of interest and allows the identification of distinct moments of interest. In our studies these
moments were observed between 150 and 450 ms after stimulus onset for facial happiness
presentation (Krolak-Salmon et al., 2006 for Rochas et al., 2013) and around 100 and 170 ms after
stimulus onset for emotional word processing (Rochas et al., 2014). Interestingly, we also could have
tested our TMS experiments with a chronological paradigm – i.e. with single pulse stimulation
launched at different timing relative to the event onset. This would also have allowed finding directly
the crucial moment for efficient disturbance. However, we used reduced intensities (80% of the
motor threshold on the left pre-SMA and 90% on the AGs) to avoid a diffusion of the induced
magnetic field and consequently to increase the precision of the stimulation in the two studies. While
this improves spatial precision, it also reduces the strength of stimulation. Single pulse TMS with
reduced intensity produces generally moderate effects on behavioural functions and would have
made it difficult to induce significant effects without a lot of repetitions. Therefore a chronological
study with TMS at different time points would have drastically increased the number of repetitions of
the stimuli and prolonged the test session. We wanted to prevent this, especially in the second study
with word materials. The electrophysiological study prior to the TMS study allowed accurate
predictions on the timing of TMS targeting to interact in specific aspects of the studied function.
67
We targeted likely important but secondary structures in the considered emotional processing.
These are of course not the only processors of the emotional functions. Other structures of a
complex network are also involved in the same function and can partly compensate for the
disruption of one of the nodes, reason for subtle behavioural effects of TMS. These networks can be
detected by scalp electrophysiology measures. In the second study, the ESI results showed significant
activity for words presented to the LVF in the left Insula and the left inferior frontal gyrus from 172 to
200ms (Fig. 4C in Rochas et al., 2014). More interestingly, interaction between the conditions –
Laterality and Emotionality – was found in brain activity in the left precuneus and in the posterior
part of the left middle temporal gyrus – Brodmann area 21. The interaction was mainly due to
stronger activity for emotional words detected in the LVF than neutral words detected in the RVF.
Hence EEG revealed potential parallel ways for the treatment of information that could have been
favoured during the functional disruption of the AGs.
The high temporal resolution of electrophysiology could indirectly enhance the resolution in space as
well. Indeed, fMRI rather favours phenomena that last for a prolonged period of time, while EEG can
also detect brief cerebral processes. The choice for a target can be done time point by time point.
This is particularly important for early periods of stimulus processing that are naturally very brief as
expressed in sharp evoked potential components. Hence even early and brief mechanisms are
accessible with these techniques.
There is a growing number of studies using a challenging combination of simultaneous TMS and EEG
(Miniussi and Thut, 2009). Such studies investigate not only where and when TMS interfere with
functions, but also how. Initial studies recorded EEG at rest in order to explore online the
propagation of the TMS effects in the brain (Ilmoniemi et al., 1997, Kähkönen et al., 2001). This
monitoring of the TMS impact on remote areas constitutes a direct measure of functional
connectivity in the brain (Schutter and van Honk, 2006; Taylor et al., 2007a). Using such studies it
could be shown that this functional connectivity could be modified by the state of the brain (Taylor et
al., 2007b; Mattavelli et al., 2013) – explaining a part of the variability of the effects of TMS in a single
subject (Romei et al., 2008). In recent examples, the index of connectivity given by concomitant EEG
and TMS was used for the evaluation of the state and the evolution of coma patients (Casali et al.,
2013; Ragazzoni et al., 2013). Apart from identifying resting state networks, the EEG-TMS
combination was also used to investigate the effects of TMS during task execution (Schürmann et al.,
2001; Thut et al., 2003). Thut and colleagues demonstrated an influence and an entrainment effect of
train of pulses on the rhythmic activity of the brain (Thut et al., 2011a). In their experiment, the short
trains of pulses entrained alpha-oscillations, though dependently on the intrinsic oscillatory activity
of the participant, influencing the subject’s level of attention. This type of investigation would have
68
been interesting in our studies as both of them used short trains of pulses. Although we had no EEG
monitoring during our TMS experiments, it is likely that similar alpha – 10 Hz – entrainment occurred
in the stimulated areas and contributed partly or totally in the observed functional effects of TMS.
The target area in our second study (the TPJ) is also classically related to attention (Corbetta and
Shulman, 2002). The addition of online EEG would have monitor the presence or not of attention
modulation phenomena. However as we will discuss in more detail below, the results of our studies
argue rather in favour of modulation of the emotional processing per se.
Electrophysiology and more particularly EEG was used in many studies on emotions and identified
several specific compounds of emotional treatment. For example, related to emotional face
processing, the generation of Mu rhythms – electrical oscillations generated in the sensory-motor
cortex in relation to actions – was found to be related to the processing of facial expression and
lateralised according to the emotional valence (Moore et al., 2012). Sel and colleagues also showed a
direct implication of the somatosensory cortex in facial emotion processing (Sel et al., 2014). Specific
implications – here related to the valence or to multisensory integration - of the somatosensory
cortex in facial emotion processing could be causally assessed using TMS and behavioural measures.
With respect to emotional words processing, evoked potential studies revealed an early posterior
negative component that was generated between 200 and 300 ms after word presentation in the left
extra-striate cortex (Kissler et al., 2006) or the fusiform gyrus (Schacht et al., 2009), both being parts
of the ventral visual stream. The disruption of this pathway with TMS could help to define the causal
role of it in emotional word processing. Generally, the abundant literature on emotional perception
in electrophysiology constitutes an important source of potential experiments with event-related
TMS. TMS could bring first a confirmation of the effective role of an area in the achievement of a
particular function, and it could also reveal subtle particularities of the studied function. Our results
showed that the coupling of event-related electrophysiology and TMS works well in a yet robust
function such as emotion. Moreover our investigations showed that TMS guided by electrophysiology
allows us to tackle very specific aspects of emotional processing.
4. The role of the premotor cortex in emotions
4.1. Motor areas for emotional recognition
The pre-SMA is narrowly associated with the motor system. It belongs to the supplementary
motor area just anterior to the primary motor cortex. It is the seat of action preparation and action
inhibition and selection in general (Rushworth et al., 2007; Mostofsky and Simmonds, 2008). In their
work on the role of the pre-SMA, Nachev and colleagues summarized the different functions related
69
to actions requiring motor programming such as “language generation, movement recognition and
ideation, maintaining working memory, establishing visuomotor associations, learning and
performing movement sequences, time perception and discrimination, “internally” guided action,
representing action intentions, conflict resolution or monitoring, and switching between action sets”
(introduction of Nachev et al., 2007). The authors argued that the pre-SMA was devoted to choose
the final action to execute among the possible motor plans and was finally a centre for the resolution
of conflicts. In an fMRI study Lau and colleagues demonstrated that the pre-SMA was also implicated
in the attention to intention of actions (Lau et al., 2004). The involvement of the pre-SMA in
emotional processing is not directly mentioned in these studies. Our investigations however lead to
conclude that it has also an important role in some emotion decoding.
Emotions are motions. The facial expressions of emotions are made of particular patterns of motions.
Although the majority of structures studied in facial emotions are not related to the motor system, a
hypothesis of movement analysis related to emotion processing seems obvious. The analysis of facial
emotions as movements adds a dimension of dynamic to the emotion issue and corresponds better
to the real life processing of emotional cue in social context for instance. In fact the processing of
dynamic facial expression impacts the performance in behavioural task and elicits larger cerebral
response compared to static emotion perception (reviewed in the introduction of Recio et al., 2014).
Some fMRI studies on dynamic facial expressions (Sato et al., 2004; Trautmann et al., 2009) reported
indeed specific additional activations first in structures apparently unrelated with the motor system
but in a subcortical regions and regions between the occipital and the temporal lobes – e.g. amygdala,
parahippocampal gyrus, inferior and middle occipital gyri, fusiform gyri, supramarginal regions, and
superior and middle temporal regions. Activations for dynamic emotional stimuli were also found in
the prefrontal cortex and even the premotor system – e.g. inferior frontal and ventral premotor
cortex, middle frontal, superior frontal areas, medial frontal and posterior cingulate cortex. Note that
all the structures showing greater activation for dynamic stimuli rather than static in an fMRI study
are not necessarily related to the direct processing of the dynamic of emotions but they could just be
entrained to react more by some of the structures that actually analyse this dynamic compound. Also
the majority of these structures, included or not in premotor and motor system, are parts of the
classical mirror neuron system – i.e. inferior frontal cortex and inferior parietal lobule (Carr et al.,
2003; Iacoboni and Dapretto, 2006). A study using a similar approach to our study found comparable
general results with slight divergences (Balconi and Bortolotti, 2013a). Authors used event-related
repetitive TMS at 10Hz over the premotor cortex during an emotion recognition task and found also
alteration of the performance of the participants for conscious or unconscious presentation.
However they mainly found effects on fear processing – effects on happiness detection were not
70
significant – with an unexpected decrease in error rate as well as reaction time – i.e. an enhancement
of the performance. The authors argued that their stimulation, longer than in our paradigms,
“activated” the stimulated area which can be doubted especially without any indication on
physiological rhythms of the area. It is somewhat difficult to believe that an arbitrary 10 Hz
stimulation activating massive population of neural cells in the target-area could mimic a relevant
physiological signal. Also strangely the figure-of-eight stimulation coil was placed just above the
midline of the brain with the maximum of induced density falling between the two hemispheres;
albeit unusual the two wings could stimulate the two premotor sides. Moreover they used only full
emotional intensities for their stimuli which could explain a part of the divergence with our results.
Although similarly, the authors claimed for a role of the premotor cortex in emotion recognition by
mirror neuron functioning. When looking at the results of our TMS study on pre-SMA (Rochas et al.,
2013) separately for each morphing level (Figure 9), although these effects were not significant, the
alteration of happiness expression recognition seemed to be more pronounced for moderate
intensities around 50 % of happiness morphing. Even if using static stimuli, the amplitude of the
observed effects and consequently of pre-SMA intervention seemed to be dependent on supposed
movement amplitude. The weak intensities of facial expression could stress for bigger need in
mimicry processing to be finely decoded and identified but they are poorly recognised anyway. The
high intensities of facial expression are prototypical signal and are easily recognised by a rather
Figure 9: unreported data from the first study of this thesis report, Rochas et al., 2013. Happiness recognition
alteration. The red curve represents the difference in performance (%) between the recognition during the vertex
stimulation minus the pre-SMA stimulation in abscissa for the different proportions of happiness in the morphing of the
presented stimuli in ordinate. The maximum of disruption of the recognition of happiness was observed for facial
expression made of 50% of happiness (not significant).
71
automatic analysis in other emotional structures – e.g. the amygdala (Breiter et al., 1996; Morris et
al., 1996; Blair et al., 1999). These exploratory results need of course to be verified in further
investigations with more participants or with other neuroimaging techniques and also with dynamic
stimuli. The use of dynamic stimuli could be particularly interesting in our experiment but it was not
really straightforward to set up within an event-related TMS paradigm. There is indeed no known
indication for a possible timing of the intervention of the premotor area in the decoding of dynamic
facial movement and consequently for the choice of the timing of stimulation according to the length
of the event. Additionally, the pre-SMA could also be responsible for judgements on intensity of
emotions without any dynamic concept. The implementation in the task of a question of judgment
on emotional intensity could also be interesting.
As already reported, a cognitive-motor group of regions including the pre-SMA and the inferior
frontal gyrus with the frontal operculum was listed as part of the neural network of emotions (Kober
et al., 2008). The authors of the review discussed the pre-SMA as responsible for selection of actions
in response to emotional cues. Evidences from intracranial electrophysiology indeed pointed to a
productive role of this specific area in emotions (Fried et al., 1998; Krolak-Salmon et al., 2006); its
electrical stimulation by DBS provoked an uncontrolled reaction and feeling of happiness in the
patients. However specific evoked responses were also recorded in the left pre-SMA for the
perception of facial happiness expression by Krolak-Salmon and colleagues (Krolak-Salmon et al.,
2006). Note that the dual implication, in production and in detection, renders a mirror functioning in
facial expression to this area, with a preference in responding to happiness. Our results of eventrelated TMS argue further in this direction, as they show an important functional role of this
premotor area in the recognition of specific facial happiness recognition. It indicates that its normal
functioning is indeed required for the correct recognition of emotions and it is not simply activated in
consequence to emotion recognition for action preparation. Our results are specific to one category
of emotion and correspond better with an analysis of facial expression than a role in action selection.
Lau and colleagues, in their study on attention to intention (Lau et al., 2004), conceded also a
functional link between the pre-SMA and the prefrontal cortex which is a major actor in emotion
regulation (Ochsner and Gross, 2005). Thus the frontal lobe does not include only a mix between
emotional processing with the prefrontal or cingulate cortex and motor processing with the motor or
premotor cortex, but it joins the two aspects for mimicry analysis of emotions.
An interesting issue with respect to the involvement of a motor-related region in the recognition of
emotions is its propensity to act either in actual emotion or just in motion. In fact, one could argue
that the pre-SMA is only responsible for a part of the analysis that is unrelated to emotions. The
analysis of movements and gestures would be indirectly involved as a generic part of any emotional
72
expression. According to our results, the specific causal role in happiness recognition argued though
in favour of an emotional processing rather than a simple movement processing. An intervention
strictly related to movement processing should have provoked alteration in any emotion recognition.
Nonetheless, it is noteworthy that facial happiness expression is maybe the emotion that implies
larger movements or movements to which pre-SMA is the most receptive. Hence the specificity of
the effects to happiness of our TMS study could be due to this specificity to induce large movements.
However the results of DBS of the pre-SMA producing reactions of uncontrolled happiness in patients
– laughter and euphoric state (Fried et al., 1998, Krolak-Salmon et al., 2006), strongly support our
interpretation of direct emotional processing in the pre-SMA.
Other neuroimaging studies on emotions found activations in the premotor cortex and the pre-SMA
which argues again for a treatment of emotions and not only a motion processing, but these
activations were not necessarily specific to happiness. These activations were found for happiness
(Seitz et al., 2008; Trautmann et al., 2009) or pleasant emotions (Hennenlotter et al., 2005) and
sometimes for other emotions (Seitz et al., 2008; Trautmann et al., 2009; also Balconi and Bortolotti,
2013a). The area seems thus not to be totally devoted to happiness. This divergence can be
explained in a distributed and integrative view of emotion processing in which a network of
structures collaborates in the recognition of the different emotions. Therefore it is not the activity of
one area that characterises one emotion but the specific pattern of activations and deactivations in a
network. A good example is provided by the study of patients with injured amygdala who showed a
bias toward happiness recognition when asked to identify facial expressions of fear or anger (Sato et
al., 2002). Moreover, the activation of one area found in classical neuroimaging studies does not
necessarily mean that it plays a crucial role for recognition. Even if an area is more or less activated
to different types of emotions the identification of the emotion could still occur in other connected
structures. Our results after TMS impairment however argue for a needed involvement of the left
pre-SMA in happiness recognition even if it can also be activated during the processing of other
emotions.
In addition, the pre-SMA shows interesting connections with emotional areas providing good
arguments for its potential early role in emotional processing. Pre-SMA and SMA are connected with
each other but the connectivity to other areas differs strongly between them. Pre-SMA does not
connect to the primary motor cortex or the spinal cord as the SMA. The pre-SMA is connected with
parts of the prefrontal cortex in particular with anterior premotor and cingulate areas as well as the
middle frontal area 46 (Luppino et al., 1993; Wang et al., 2005). The pre-SMA is thus connected with
the orbitofrontal cortex that reacts early to emotional stimuli (Kawasaki et al., 2001). Moreover it
receives projections from the different sensory modalities (Rolls, 2004) and from the thalamus as
73
described in monkeys (Inase et al., 1996). These connected areas confer potential precocious
involvement in emotional processing to the pre-SMA.
4.2. Alteration of the mimicry system
The effects observed during the stimulation of the pre-SMA (Rochas et al., 2013) supported
the hypothesis of an intervention of mirror mechanisms in the recognition of facial expression of
emotions. Part of these mirror mechanisms were presumed to involve the premotor system and our
experiment tested such implication of the pre-SMA. Although classically implicated in motor
processing and action regulation as already discussed above, the area is closely related to mirror
neuron system as part of the premotor cortex which is also often related to the processing of
emotions, and notably the facial expressions of emotions. The premotor cortex of the macaque
monkey was the first area showing mirror response in intracortical cell recording (di Pellegrino et al.,
1992) – i.e. it contains neurons responding while the monkey performs an action and while it
observes a similar action done for instance by the experimenter. Before the clear evidence of a
mirror neuron system, analogous mirror neuron mechanisms were already found in EEG signals with
the suppression of Mu rhythms during both execution and observation of actions (Cohen-Seat et al.,
1954; Gastaut and Bert, 1954). Later, an MEG study analysing electrophysiological rhythms also
showed similar suppression of ulnar stimulation evoked rebound in oscillations around 20 Hz in the
precentral region during object manipulation and action observation (Hari et al., 1998). Some
evidences also came from studies of the motor evoked potential in response to TMS over the motor
cortex that was modulated by the observation of an action involving the muscle corresponding to the
specifically stimulated motor cortex field (Fadiga et al., 1995). Former macaque monkey results were
confirmed in left inferior frontal cortex using fMRI in humans and complemented with a second
region showing similar activation pattern and located in the right anterior parietal cortex (Iacoboni et
al., 1999). The mirror neuron system was then studied in various aspects of human behaviour
through action perception and understanding, and in motor learning (Rizzolatti et al., 2001).
Mimicry mechanisms are also suspected to act extensively in emotion processing and in particular in
facial expression decoding. Dimberg and colleagues found that the presentation of emotional facial
expressions provoked automatic contractions of congruent facial muscles in the observer (Dimberg,
1982). Similar rapid and automatic facial mimicry were then described in various contexts (Hartfield
et al., 1993; Vrana and Gross, 2004; Sato and Yoshikawa, 2007). The neural substrates underlying the
process of facial mimicry – related to emotion expression-perception – were investigated in fMRI and
included the classic areas of the mirror neuron system with the inferior frontal cortex and the inferior
74
partial lobule in relation with the superior temporal cortex, but also in addition the insula and
amygdala showed similar response (Carr et al., 2003; see also Iacoboni and Dapretto, 2006).
Interestingly the fMRI activity of some of the areas showed variability in their responses according to
the type of emotion (van der Gaag et al., 2007) and the dynamic nature of the stimuli (Sato et al.,
2004). Notably, the emotional processing has been shown to modulate the mirror neuron system
excitability in a TMS experiment (Hill et al., 2013); the motor evoked potential by TMS over the
primary motor cortex during action observation were modulated according to preceding emotional
stimuli. In another TMS study, the early (< 200 ms after stimuli) disruption of a right somatosensory
cortex altered the performance of participants in an emotional matching task (Pourtois et al., 2004).
This result suggested a participation of a sensory cortex in mimicry mechanisms for emotional
perception. Numerous neuroimaging studies showed additional areas in interaction with the mirror
neuron system in the processing of emotional stimuli. The perception of pleasant emotion
expressions engaged mirror neuron response as well in the motor, somatosensory and limbic
systems (Hennenlotter et al., 2005). The pre-SMA in this last study was implicated in the motor
execution. In another study, the SMA showed correlated activity with the muscular response in facial
expression mimicry (Likowski et al., 2012). Furthermore mirror neuron activity was found in the
proper pre-SMA although it responded more to execution than observation of facial emotional
expressions or hand-grasping actions as shown in direct response recording in intracranial recording
(Mukamel et al., 2010) and the area showed also indirect response to gesture perception (Villarreal
et al., 2008). The pre-SMA was also implicated in emotional mimicry processes in a deep brain
electrode study in a patient that initiated the present work (Krolak-Salmon et al., 2006). However the
contribution of the mirror neuron system in the achievement of emotion recognition needed further
validation. According to our study on facial expression recognition, the mimicry mechanisms
observed in the pre-SMA were actually responsible for emotional expression recognition (Rochas et
al., 2013). Consequently the pre-SMA may act as a mirror neuron area in the decoding of emotional
gesture. Interestingly the major part of the network responding to dynamic emotional stimuli found
in two fMRI studies (Sato et al., 2004; Trautmann et al., 2009) is intermixed with the mirror neuron
system. Taking into account its double nature – movement and emotional processing – the pre-SMA
constitutes an interesting link between emotion and movement, nicely supporting the etymological
relation between the words emotion and motion. Two event-related TMS studies showed alteration
of facial expression recognition by stimulating the premotor cortex after the emotional stimuli
(Balconi and Bortolotti, 2013a) or before the stimuli onset (Balconi and Bortolotti, 2013b). The
authors claimed that the effects were related to mirror mechanisms. The deep brain stimulation
studies (Fried et al., 1998; Krolak-Salmon et al., 2006) and our conclusions (Rochas et al., 2013)
tended to show that the left pre-SMA has a specific role in processing happiness expression and
75
maybe more specifically by the decoding movement of prototypic happy face movement – i.e. a
smile. These results could suggest a possible preference of the area for zygomaticus mimicry.
Experiments using cut pictures of face parts could answer this issue on facial feature specificity.
4.3. A prefrontal question
As discussed previously (II.5.2.), two distinct neuronal systems are involved in empathy and
mimicry (van der Gaag et al., 2007); the prefrontal cortex and the other classic areas for empathic
appropriation of others’ feelings on one side, and the premotor and motor cortex for the recognition
of emotional movements and actions by mimicry on the other side. The association of emotion
recognition and emotion regulation or integration refers therefore to the bases of interoception and
somatic markers stated in the hypothesis of Damasio on emotional processing from the inner signals
(Damasio, 1996). The two entities can play complementary roles in the recognition of emotions in
others and in the reaction to this recognition. The premotor and motor systems can participate in the
recognition of perceived facial expressions and their dynamic evolution. The pre-SMA is involved in a
network of dynamic facial expression recognition (Trautmann et al., 2009). This point is corroborated
by our TMS results in which happy face expression is impaired in certain specific intensities of
expression (not significant, Fig. 9). Hence the mirror neuron system by embodiment – i.e. the
incarnation of emotions filtered in the one’s own body feelings – can be considered as a system of
interpretation and appropriation of other’s emotional expressions by automatic mimicry. This
acquired information is integrated and participate in emotion regulation through the amygdala,
cingulate or prefrontal cortex. The traditional network of emotions can also be responsible for the
feelings in reaction to the recognition of an emotional state or situation. Then the premotor cortex in
parallel could act for preparation of a productive reaction or a behavioural response (Kober et al.,
2008) in tandem with the motor cortex. Additionally, the pre-SMA being known for its implication in
intention processing (Lau et al., 2004), it may act in early steps of facial emotion decoding in order to
interpret the intention of others and prepare a response. Even if their argumentation was not
directed towards emotional processing, Rushworth and colleagues in a review work also discussed
the idea of an acquisition of information about other individuals in the anterior cingulate cortex,
belonging to the prefrontal cortex and limbic system (Rushworth et al., 2007). The premotor system
would also analysis strength and types of the others’ emotional signals in order to qualify their
intention. On the other hand, the amygdala and coprocessors (insula, cingulate cortex) can then
evaluate one’s own response and generate appropriate feelings towards it. Therefore these two
systems would act in parallel. Interestingly, the DBS of left pre-SMA provoked unrestrained reaction
of happiness and joy feeling (Fried et al., 1998; Krolak-Salmon et al., 2006). The mood of the patients
76
was thus modified. These results suggest an implication of the pre-SMA in motor program of
happiness but in interaction with affective state – possibly by connexions with the cingulate cortex.
One could ask if it is mimicry for empathy. The two systems seem to be in interaction and the preSMA is one of their intersections. Note that our paradigm of brief event-related TMS over the preSMA was lacking of long-term effects and did not show any alteration of the affective state.
Consequently we did not observe any changes in the mood auto-evaluation between before and
after the TMS procedure. One can expect that repetitive paradigm of stimulation of the pre-SMA
could act on mood regulation and it would be an interesting eventual application in patients with
mood disorders. Finally as a close part of the mirror neuron system and an actor in mimicry, the preSMA placed this work in a social dimension. The process studied is a part of the interaction between
individuals.
Concerning the question of emotional lateralisation, even at the frontal lobe level, our results after
disruption of the left pre-SMA argues in favour of valence hypothesis rather than approachwithdrawal model (see II.5.1). Indeed the effects on emotion recognition are observed for happiness
only, and not for anger which is an approach emotion. Positive emotion is in this case processed in
the left hemisphere. Our results go also against the right hemisphere hypothesis for dominance of
emotion processing at least in the pre-motor cortex. However in view of a distributed model, it
would be just specific to happiness, without consideration of any emotional dimension.
4.4. A functional model around the target
Based on our results and the literature, I here present a model of the putative functional
relation between the major actors in facial emotion recognition in relation with the left pre-SMA (Fig.
10). The visual information may arrive in the pre-SMA directly from a thalamic relay. Also the preSMA has connections with parts of the prefrontal, cingulate or orbitofrontal cortex which could
provide a signal flux from sensory modalities, possibly already tinted with an emotional colour. The
classic emotional network (in yellow) could have a bidirectional relation with the motor system (in
green) for emotional signal decoding and emotional regulation. The pre-SMA with mirror neuron
properties acts in movement analysis and may interact with the visual pathway and notably in middle
temporal area for motion perception. Also due to its intervention in facial expression, a relation with
the fusiform face area is expected. In cooperation with the emotional network, the motor system
underlies facial emotion processing in relation with movement decoding. This additional processing
refines possibly the perception of emotional expressions.
77
Figure 10: Facial expression recognition model around the pre-SMA. Arrows represent the information transfer. Grey
arrows represent the processing potentially altered by TMS over the left pre-SMA.
4.5. Implications
Beyond the simple modification of the perception of emotional stimuli in a given behavioural
task, our investigation could open the doors for more general modification. Considering the first
presented work, one could extract a concept of driving in the manipulation of emotional perception
by acting on the pre-SMA. Emotional state and thus emotional perceptions which modify this state,
have a clear influence on behaviour as attention to little stimuli or interpretation of complex ideas.
The pre-SMA is a key area in action selection and would have a primordial role in execution of actions
and the interaction with others. Hence one could speculate the potential of our TMS manipulation on
the free will and behaviour of the individual while both its emotional state and its action execution
are potentially modified.
As a keystone between motion and perception, the role of mimicry is naturally related to social
science or neuropathology domains. These results could also help to understand and find strategies
in the treatment of pathologies in relation with mirror neuron system such as autism. Also the effect
78
of mood modulation observed in DBS of the left pre-SMA can be considered as a potential
therapeutic approach in many mood disorder diseases.
5. The right hemisphere an early reader of emotions
The second presented work (Rochas et al., 2014) was devoted to the investigation of the role
of the right hemisphere both in a language function – written word reading – and in the processing of
emotional stimuli – emotional words. Derived from the right hemisphere hypothesis, our experiment
studied rather the possible interaction or cooperation between the two hemispheres in reading and
the place of the emotional words in this process. We studied two main questions with respect to
emotional word processing. First we were interested in the early processing of emotionality in word
material in the right hemisphere, known to be dominant for emotions. Secondly we wanted to see if
the detection of emotional words was possible in the right hemisphere, being non-dominant for
language processing.
5.1. Processing of emotions in the right hemisphere
We will not return in details on the rationale of a dominance of the right hemisphere in
emotional processing (see section II.5.1.). In numerous studies on emotions in language, it was
suggested that the right hemisphere plays a dominant role (reviewed in Landis, 2006; Landis et al.
1982; Landis et al. 1983; Reuterskiöld 1991; Borod et al. 1992; Lalande et al. 1992; Cicero et al. 1999;
Graves et al., 1981; Ortigue et al., 2004; Smith and Bulman-Fleming, 2005; Mohr et al., 2005;
Frühholz et al. 2011; Ponz et al. 2013). However its supremacy in emotional language processing is
also debated. In our study, the role of the right hemisphere was tested in a written language task
through a contrast between neutral and emotional words (Rochas et al., 2014). In order to identify a
possible target serving the modulation of the perception of emotion by TMS, the contrast was first
revealed using EEG. The analysis of the electrical source imaging at early latencies according to the
factor Emotionality allowed the identification of specific processors of emotion in word reading; a
particular hotspot was identified in the right hemisphere, the temporoparietal junction (TPJ).
Some authors proposed that emotional materials induce a potentiation of the activations in the
regular language networks (Citron, 2012; Schacht and Sommer, 2009; Palazova et al., 2013; Keuper et
al., 2014). In our study (Rochas et al., 2014), we rather proposed the implication of supplementary
structures, more specialised in the emotional aspects of words. The use of a simultaneous bilateral
presentation in the task made it possible to compare the initial activation independently in each
hemisphere. The differences due to the factor Emotionality were not expected in the classical left
79
hemisphere network for language. Our study confirmed the hypothesis of processors dedicated to
emotionality of word in addition, and in interplay with the regular structures of word analysis. Our
results are in accordance with the emotional theory of the dominance of the right hemisphere in
language.
Our analysis strategies focused on early time of processing and went thereby against the idea of an
exclusively late intervention of the right hemisphere in emotional processing of word material
(Abbassi et al., 2011) – i.e. after the left hemisphere. Conversely, our results argued indeed in favour
of early steps of treatment of emotionality of words in the right hemisphere. Such early processing
differences may have been possibly underestimated with other functional neuroimaging techniques
– i.e. fMRI. This treatment of emotionality in the right hemisphere was apparently even present for
words from the RVF – right TPJ activation for Emotional factor alone in EEG (Fig. 4B in Rochas et al.,
2014) and interaction in TMS results (Fig. 5 in Rochas et al., 2014) – and suggests thus a possible
direct access of the right hemisphere for emotional information coming from the RVF. This direct
ipsilateral transfer of a part of the visual signal is maybe possible by a mix of fibres at the level of the
optical chiasm or at the primary visual cortex between the two hemispheres. A transfer from the left
AG to the right is rather unlikely but cannot be excluded. One can also think that a transfer might
occur via subcortical structures. Further investigations on this early activation of the right
hemisphere are needed in order to characterise this transfer. Chronologic single pulse TMS on the
one or other AG could discard a transmission from the left to the right AG for instance. Also TMS over
the occipital visual cortices could be applied in order to investigate the role of the two sides in the
information transfer.
As already discussed, the non-dissociation of the effects of stimulation between the two AG suggests
an interaction between the two homologous areas or, a transfer of signals from the right AG to the
left hemisphere via the left AG. The left hemisphere could thus not be totally excluded from
emotional word processing. Moreover the significant interaction in the ESI results of the EEG study
(Rochas et al., 2014) argues also in this direction with activations in the left precuneus and in the
posterior part of the left middle temporal gyrus.
Due to a well know implications of the TPJ and especially the right one in attention regulation, our
observations could be suspected to result from the induction of an attentional bias rather than a real
modulation of perception of emotional words. Indeed the right temporoparietal cortex is known to
prepare the detection of relevant stimuli, particularly when they are salient and unexpected
(Corbetta and Schulman, 2002). Nevertheless the effects that we observed in both EEG and TMS
were always related to the emotional factor. It thus seems not to be just a disruption of attention but
an emotion- and word-related phenomenon.
80
5.2. Pre-processor of written material in the right hemisphere
Is the right hemisphere able to read? The specific activation found in EEG in the right TPJ for
the emotional words proved to be necessary to the normal functioning of word detection in the
event-related TMS study. The divergence of implication of this area between neutral and emotional
word detection implies also a lexical characterisation of the words. In fact to be differentiated
between neutral or emotional, letter strings need to be recognised as words. The temporal and
parietal lobes and more particularly the temporoparietal junctions (TPJ) are implicated in various
aspects of language processing (for review see Price, 2012). This network is though largely lateralised
in the left hemisphere even if as discussed in the article, bilateral involvements are also possible
(Kuchinke et al., 2005; Graves et al., 2010; Vigneau et al., 2011; Diaz and Hogstrom, 2011). The early
implication claimed in our results argues in favour of an actual recognition of words in the right
hemisphere. Using a lexical decision task our investigations were oriented on comprehension
processes of language and not on production. It is known that the lateralisation of the neural
substrates for language is particularly verified for the productive functions and is much less clear for
the comprehension part, especially in the auditory modality (Lidzba et al., 2011). Hence our
comprehension task facilitated maybe a bilateral involvement. Similar investigations on a language
production task would be therefore interesting.
Our experiments focused on the early steps of the processing – around the P100 component – in
order to identify precocious lateralised mechanisms of primary access to the meaning of words. The
primary visual cortex is connected to the precuneus which is also connected to the TPJ. The
precuneus constitutes a possible relay and included already raw activations and differences in
activation according to the factor Laterality in ESI results (Fig. 4A in Rochas et al., 2014). This pathway
can convey first steps of lexical analysis. Note that the later stages of cerebral processing are possibly
more mixed between the two hemispheres. The chronology of word processing was accessible
thanks to the temporal resolution of the electromagnetic imaging and it was possibly verified with
event-related TMS. Conversely the use of rTMS could have induced remote effects in areas
connected to the target-area by direct entrainment or by subsequent plasticity mechanisms.
Moreover the time-locked stimulation limits the disruptive effects to early processing expected in the
temporoparietal junctions.
As discussed in the previous part, the early specific treatment of emotional words in the right
hemisphere contradicts an exclusively late implication of the right hemisphere (Abbassi et al., 2011)
or a simply specific activation and treatment of emotional words in the though early but classic
81
pathway for written words as argued by other authors (Citron, 2012; Schacht and Sommer, 2009;
Palazova et al., 2013; Keuper et al., 2014). Nonetheless, our results argue for a possible parallel
analysis of written words in the right hemisphere independently of the visual word form area located
in the left lateral occipitotemporal sulcus (Dehaene and Cohen, 2011). Using similar paradigms, the
degree of intervention of our right hemisphere’s area could be tested in more details with various
categories of words and non words to assess its role according to emotionality, emotional valence,
words length, or familiarity for instance. Also it is possible that the paradigm of presentation forced
the two hemispheres to work in word decoding in abnormal proportions compared to a normal
situation – i.e. with central presentation. Indeed the simultaneous presentation of letter strings in
both hemifields provoked certainly an interaction of primary analysis of letters even when they are
pseudo-words. We might have used isolated words – with a shorter display to keep the same level of
difficulty in lexical detection – in separated hemifield or in the centre of the visual field. The direct
access of information to the right TPJ discussed in the previous part could also serves the reading
analysis by decoding emotional cue. Left and right sides cooperate certainly in reading process.
The fact that the effects of event-related TMS were a slowing of the responses and not a decrease of
performances indicates either that the disrupted area process the information slower after
stimulation or more probably that the TMS disrupted completely the area and the detection of the
emotionality information is overtaken by the rest of the networks. This network could be in the left
dominant hemisphere for language – excluding the left AG itself which is targeted by the TMS – or in
classic emotional structures – e.g. amygdala. Interestingly, in addition to the factor-dependent
effects an interaction in ESI results between the conditions of presented words – Laterality and
Emotionality – was found in left hemisphere – i.e. left precuneus and posterior part of the left middle
temporal gyrus – Brodmann area 21. The interaction effect related stronger activities for emotional
words detected in the LVF compared to neutral words in the RVF. Thus these left hemisphere
structures revealed in EEG could have overtaken detection of RVF emotional words during the
disruption of AG by TMS. However the slowing effect of TMS suggest that this secondary network
when deprived of important processors, notably the TPJs, lead to a slower processing due to a non
specialisation – for emotionality – or to a larger number of steps to receive information – for word
coming from the LVF by the right hemisphere. The detection is performed but it is slower.
5.3. A functional model around the targets
Based on the event-related EEG and TMS results, precocious steps of the processing of
emotional words in the left or right visual field are proposed and summarised in a schematic model
82
(Fig. 11). The visual information from each hemifield arrives independently in each primary visual
cortex and a first lexical analysis is initiated with the transfer to the precuneus. This left structure
showed significant activation for the detection of word in the left visual field compared to the right
visual field. The words are differentiated from pseudo-words. As emotionality analysis has been
shown to be related to the activation of the right TPJ we hypothesise a transfer of information to this
structure. This transfer could be done from the right precuneus for words in the LVF or from the left
TPJ for the words from the RVF. Hypothetical transfers between the two primary visual cortices or
the precuneus are not represented as well as a direct transfer from thalamic or amygdala structures.
All this steps occur in around 200 ms. After the initiation of the emotional processing, the words are
sent to the left language network for further processing.
Figure 11: Functional model of emotional word detection in hemifield. Arrows represent the information transfer. Grey
arrows represent the processing potentially altered by TMS over the right TPJ.
83
VIII. CONCLUSION
Thanks to electrophysiology we were able to target cortical areas in really specific functions
of emotional perception. As expected the left pre-SMA play an important role in the processing of
facial happiness expression (Rochas et al., 2013) and the TPJ are implicated in the processing of
words according to their emotionality and their side of presentation (Rochas et al., 2014). Moreover
the TMS disruption of the target area evidenced important aspects of the studied emotional
processing. The implication of pre-SMA in happiness recognition seems to be intensity dependent
and implies certainly a motor decoding mechanism. The intervention of the right TPJ in emotional
word detection is part of an interaction and cooperation between the two hemispheres in word
detection. Comparing the two studies, the modifications induced by TMS were restricted to
categories of stimuli. The moderate intensities of facial happiness recognition were altered in the
first study while the perception of emotional words in the LVF was slowed in the second study. The
modifications were limited – either accuracy or reaction times – but also specific. The specificity of
the modifications allows for potential targeted strategies of modification in future studies.
IX. REFERENCES
A
Abbassi E, Kahlaoui K, Wilson MA, Joanette Y. 2011. Processing the emotions in words: the complementary contributions of
the left and right hemispheres. Cogn Affect Behav Neurosci, 11(3): 372-85.
Adolphs R, Gosselin F, Buchanan TW, Tranel D, Schyns P, Damasio AR. 2005. A mechanism for impaired fear recognition
after amygdala damage. Nature, 433(7021): 68-72.
Adolphs R, Tranel D, Damasio AR. 2003. Dissociable neural systems for recognizing emotions. Brain Cogn, 52(1): 61-9.
Afra J, Mascia A, Gerard P, Maertens de Noordhout A, Schoenen J. Interictal cortical excitability in migraine - a study using
transcranial magnetic stimulation of motor and visual cortices. Ann Neurol 1998; 44:209-215.
Ahdab R, Ayache SS, Brugières P, Goujon C, Lefaucheur JP. 2010. Comparison of "standard" and "navigated" procedures of
TMS coil positioning over motor, premotor and prefrontal targets in patients with chronic pain and depression.
Neurophysiol Clin. 40(1):27-36.
Alagona G, Bella R, Ferri R, Carnemolla A, Pappalardo A, Costanzo E, Pennisi G. 2001. Transcranial magnetic stimulation in
Alzheimer disease: motor cortex excitability and cognitive severity. Neuroscience Letters, 314(1–2): 57–60.
Alemà G, Donini G. 1960. Sulle modificazioni cliniche ed elettroencefalografiche da introduzione intracarotidea di iso-amiletil-barbiturato di sodio nell’uomo. Bollettino della Società Italiana di Biologia Sperimentale, 36: 900–904.
Aleman A. 2013. Use of Repetitive Transcranial Magnetic Stimulation for Treatment in Psychiatry. Clin Psychopharmacol
Neurosci, 11(2):53-59.
84
Amassian VE, Eberle L, Maccabee PJ, Cracco RQ. Modelling magnetic coil excitation of human cerebral cortex with a
peripheral nerve immersed in a brain-shaped volume conductor: the significance of fiber bending in excitation.
Electroenceph clin Neurophysiol 1992; 85:291-301.
Arana AB, Borckardt JJ, Ricci R, Anderson B, Li X, Linder KJ, Long J, Sackeim HA, George MS. 2008. Focal electrical stimulation
as a sham control for repetitive transcranial magnetic stimulation: Does it truly mimic the cutaneous sensation and pain of
active prefrontal repetitive transcranial magnetic stimulation? Brain Stimul. 1(1):44-51.
Atchley RA, Stringer R, Mathias E, Ilardi SS, Minatrea AD. 2007. The right hemisphere's contribution to emotional word
processing in currently depressed, remitted depressed, and never-depressed individuals. Journal of Neurolinguistics, 20(2):
145–160.
Attarian S, Azulay JP, Lardillier D, Verschueren A, Pouget J. 2005. Transcranial magnetic stimulation in lower motor neuron
diseases. Clin Neurophysiol, 116(1): 35-42.
Aurora SK, Ahmad BK, Welch KMA, Bhardhwaj P, Ramadan NM. Transcranial magnetic stimulation confirms
hyperexcitability of occipital cortex in migraine. Neurology 1998; 50:1111-1114.
Avery DH, Holtzheimer PE 3rd, Fawaz W, Russo J, Neumaier J, Dunner DL, Haynor DR, Claypoole KH, Wajdik C, Roy-Byrne P.
2007. Transcranial magnetic stimulation reduces pain in patients with major depression: a sham-controlled study. J Nerv
Ment Dis, 195(5): 378-81.
B
Bae EH, Schrader LM, Machii K, Alonso-Alonso M, Riviello JJ Jr, Pascual-Leone A, Rotenberg A. 2007. Safety and tolerability
of repetitive transcranial magnetic stimulation in patients with epilepsy: a review of the literature. Epilepsy Behav, 10(4):
521-8.
Bae EH, Theodore WH, Fregni F, Cantello R, Pascual-Leone A, Rotenberg A. 2011. An estimate of placebo effect of repetitive
transcranial magnetic stimulation in epilepsy. Epilepsy Behav, 20(2):355-9.
Bae JS, Simon NG, Menon P, Vucic S, Kiernan MC. 2013. The Puzzling Case of Hyperexcitability in Amyotrophic Lateral
Sclerosis. J Clin Neurol, 9(2): 65-74.
Baeken C, De Raedt R, Bossuyt A, Van Hove C, Mertens J, Dobbeleir A, Blanckaert P, Goethals I. 2010. The impact of HFrTMS treatment on serotonin(2A) receptors in unipolar melancholic depression. Brain Stimul, 4(2): 104-11.
Balconi M, Ferrari C. 2012. rTMS stimulation on left DLPFC increases the correct recognition of memories for emotional
target and distractor words. Cogn Affect Behav Neurosci, 12(3): 589-98.
Balconi M, Bortolotti A. 2013a. Conscious and unconscious face recognition is improved by high-frequency rTMS on premotor cortex. Conscious Cogn, 22(3): 771-8.
Balconi M, Bortolotti A. 2013b. The "simulation" of the facial expression of emotions in case of short and long stimulus
duration. The effect of pre-motor cortex inhibition by rTMS. Brain Cogn, 83(1): 114-20.
Bard P, Rioch DM. 1937. A study of four cats deprived of neocortex and additional portions of the forebrain. John Hopkins
Med J. 60: 73–153.
Bard P. 1928. A diencephalic mechanism for the expression of rage with special reference to the central nervous system.
Am J Physiol. 84: 490–513.
Barker AT, Freeston IL. 1985. Medical applications of electric and magnetic fields. Electron. Power 31(10). 757-60.
Barker AT, Freeston IL, Jalinous R, Jarratt JA. (1987) Magnetic stimulation of the human brain and peripheral nervous
system: an introduction and the results of an initial clinical evaluation. Neurosurgery; 20:100±109.
Barker AT, Freeston IL, Jalinous R, Merton PA, Morton HB (1985): Magnetic stimulation of the human brain. J. Physiol.
(Lond.) 369: 3P.
85
Barker AT, Garnham CW, Freeston IL. 1991. Magnetic nerve stimulation: the effect of waveform on efficiency,
determination of neural membrane time constants and the measurement of stimulator output. Electroencephalogr Clin
Neurophysiol Suppl. 43:227-37.
Barker AT, Jalinous R, Freeston IL. 1985. Non-invasive magnetic stimulation of human motor cortex. Lancet, 1(8437): 1106-7.
Barker AT. 1991. An introduction to the basic principles of magnetic nerve stimulation. J Clin Neurophysiol, 8:26–37.
Barker AT. 2002. The history and basic principles of magnetic nerve stimulation. In: Pascual-Leone A, Davey N, Rothwell J,
Wasserman E, Puri B, eds. Handbook of transcranial magnetic stimulation. Arnold: 3–17. London.
Barrett LF. 2006. Are emotions natural kinds? Perspectives on Psychological Science, 1 : 28-58.
Bartholow R. 1874. Experimental investigations into the functions of the human brain. The American Journal of the Medical
Sciences, 134: 305–313.
Batty M, Taylor MJ. 2003. Early processing of the six basic facial emotional expressions. Brain Res Cogn Brain Res, 17(3):
613-20.
Baumert A, Sinclair C, MacLeod C, Hammond G. 2011. Negative emotional processing induced by spoken scenarios
modulates corticospinal excitability. Cogn Affect Behav Neurosci, 11(3):404-12.
Bechara A, Tranel D, Damasio H, Adolphs R. 1995. Double dissociation of conditioning and declarative knowledge relative to
the amygdala and hippocampus in humans. Science, 269: 1115–1118.
Bediou B, Krolak-Salmon P, Saoud M, Henaff MA, Burt M, Dalery J, d’Amato T. 2005a. Facial expression and sex recognition
in schizo- phrenia and depression. Can J Psychiatry. 50(9): 525–33.
Bediou B, Franck N, Saoud M, Baudouin JY, Tiberghien G, Daléry J, d'Amato T. 2005b. Effects of emotion and identity on
facial affect processing in schizophrenia. Psychiatry Res, 133: 149–157.
Bediou B, Asri F, Brunelin J, Krolak-Salmon P, D'Amato T, Saoud M, Tazi I. 2007. Emotion recognition and genetic
vulnerability to schizophrenia. Br. J. Psychiatry, 191: 126-30.
Berardelli A, Rona S, Inghilleri M, Manfredi M. 1996. Cortical inhibition in Parkinson's disease. A study with paired magnetic
stimulation. Brain, 119(Pt1): 71-7.
Bickford RG, Fremming BD (1965). Neuronal stimulation by pulsed magnetic fields in animals and man. Digest of the 6th
International Conference of Medical Electronics and Biological Engineering, p. 112.
Bisiacchi PS, Cona G, Schiff S, Basso D. 2011. Modulation of a fronto-parietal network in event-based prospective memory:
An rTMS study. Neuropsychologia, 49(8): 2225–2232.
Blair RJ, Morris JS, Frith CD, Perrett DI, Dolan RJ. 1999. Dissociable neural responses to facial expressions of sadness and
anger. Brain, 122(5): 883-93.
Blanchette I, Richards A. 2013. Is emotional Stroop interference linked to affective responses? Evidence from skin
conductance and facial electromyography. Emotion, 13(1): 129-38.
Blasi G, Hariri AR, Alce G, Taurisano P, Sambataro F, Das S, Bertolino A, Weinberger DR, Mattay VS. 2009. Preferential
amygdala reactivity to the negative assessment of neutral faces. Biol Psychiatry, 66(9): 847-53.
Borckardt JJ, Reeves ST, Frohman H, Madan A, Jensen MP, Patterson D, Barth K, Smith AR, Gracely R, George MS. 2011. Fast
left prefrontal rTMS acutely suppresses analgesic effects of perceived controllability on the emotional component of pain
experience. Pain, 152(1): 182-7.
Borckardt JJ, Smith AR, Reeves ST, Madan A, Shelley N, Branham R, Nahas Z, George MS. 2009. A pilot study investigating
the effects of fast left prefrontal rTMS on chronic neuropathic pain. Pain Med, 10(5):840-9.
86
Borckardt JJ, Smith AR, Reeves ST, Weinstein M, Kozel FA, Nahas Z, Shelley N, Branham RK, Thomas KJ, George MS. 2007.
Fifteen minutes of left prefrontal repetitive transcranial magnetic stimulation acutely increases thermal pain thresholds in
healthy adults. Pain Res Manag, 12(4): 287–290.
Bordier H. 1897. Précis d'électrothérapie : galvanisation, voltaïsation sinusoïdale, faradisation, franklinisation, haute
fréquence, électrophysiologie, électrodiagnostic et électrothérapie proprement dite. Paris. J.-B. Baillière et fils.
Borod JC, Koff E, Lorch MP, Nicholas M. 1988. Emotional and non-emotional facial behaviour in patients with unilateral
brain damage. Journal of Neurology, Neurosurgery and Psychiatry, 51(6) : 826–832.
Borod JC, Andelman F, Obler LK, Tweedy JR,Welkowitz J. 1992. Right hemisphere specialization for the identification of
emotional words and sentences: evidence from stroke patients. Neuropsychologia, 30:827-844.
Borod JC, Bloom R, Haywood CS. 1998. Verbal aspects of emotional communication. M. Beeman, C. Chiarello (Eds.), Right
hemisphere language comprehension: Perspectives from cognitive neurosciences, Erlbaum, Mahwah, NJ.
Boroojerdi B, Meister IG, Foltys H, Sparing R, Cohen LG, Töpper R (2002) Visual and motor cortex excitability: a transcranial
magnetic stimulation study. Clin Neurophysiol 113: 1501–1504.
Breiter HC, Etcoff NL, Whalen PJ, Kennedy WA, Rauch SL, Buckner RL, Strauss MM, Hyman SE, Rosen BR. 1996. Response
and habituation of the human amygdala during visual processing of facial expression. Neuron, 17(5): 875-87.
Brigo F, Storti M, Tezzon F, Manganotti P, Nardone R. 2013. Primary visual cortex excitability in migraine: a systematic
review with meta-analysis. Neurol Sci, 34(6): 819-30.
Brix G, Seebass M, Hellwig G, Griebel J. Estimation of heat transfer and temperature rise in partial-body regions during MR
procedures: an analytical approach with respect to safety considerations. Magn Reson Imag 2002;20:65–76.
Brunelin J, Poulet E, Bediou B, Kallel L, Dalery J, D'Amato T, Saoud M. 2006. Low frequency repetitive transcranial magnetic
stimulation improves source monitoring deficit in hallucinating patients with schizophrenia. Schizophr Res, 81(1): 41–45.
Brunet D, Murray MM, Michel CM. 2011. Spatiotemporal analysis of multichannel EEG: CARTOOL. Comput Intell Neurosci,
2011:813870.
Buelte D, Meister IG, Staedtgen M, Dambeck N, Sparing R, Grefkes C, Boroojerdi B. 2008. The role of the anterior
intraparietal sulcus in crossmodal processing of object features in humans: An rTMS study. Brain Research, 1217: 110–118.
C
Cabanac M. 2002. What is emotion? Behavioural Processes, 60(2): 69–83.
Cacioppo JT, Petty RE, Losch ME, Kim HS. 1986. Electromyographic activity over facial muscle regions can differentiate the
valence and intensity of affective reactions. J Pers Soc Psychol, 50(2): 260-8.
Cahill L, Haier RJ, Fallon J, Alkire MT, Tang C, Keator D, Wu J, McGaugh JL. 1996. Amygdala activity at encoding correlated
with long-term, free recall of emotional information. Proc Natl Acad Sci USA, 93(15): 8016-21.
Calvo MG, Marrero H, Beltrán D. 2013. When does the brain distinguish between genuine and ambiguous smiles? An ERP
study. Brain Cogn, 81(2): 237-46.
Campana, G., Cowey, A., Walsh, V. 2003. Priming of motion direction and area V5/MT: a test of perceptual memory. Cereb.
Cortex 12, 663–669
Camprodon JA, Martínez-Raga J, Alonso-Alonso M, Shih MC, Pascual-Leone A. 2007. One session of high frequency
repetitive transcranial magnetic stimulation (rTMS) to the right prefrontal cortex transiently reduces cocaine craving. Drug
and Alcohol Dependence, 86(1): 91–94.
Cannon WB. 1927. The James–Lange theory of emotions: a critical examination and an alternation. Am J Psychol, 39: 106–
124.
87
Cantello R, Gianelli M, Bettucci D, Civardi C, De Angelis MS, Mutani R. 1991. Parkinson's disease rigidity: magnetic motor
evoked potentials in a small hand muscle. Neurology, 41(9): 1449-56.
Carbunaru R, Durand DM. 2001. Toroidal coil models for transcutaneous magnetic stimulation of nerves. IEEE Transactions
on Biomedical Engineering, 48: 434–441.
Carr L, Iacoboni M, Dubeau MC, Mazziotta JC, Lenzi GL. 2003. Neural mechanisms of empathy in humans: a relay from
neural systems for imitation to limbic areas. Proc Natl Acad Sci USA, 100(9): 5497-502.
Casali AG, Gosseries O, Rosanova M, Boly M, Sarasso S, Casali KR, Casarotto S, Bruno M-A, Laureys S, Tononi G, Massimini M.
(2013) A theoretically based index of consciousness independent of sensory processing and behavior. Science translational
medicine, 5(198): 198ra105.
Chammat M, Foucher A, Nadel J, Dubal S. 2010. Reading sadness beyond human faces. Brain Res. 1348: 95-104.
Chen R, Classen J, Gerloff C, Celnik P, Wassermann EM, Hallett M, Cohen LG. 1997. Depression of motor cortex excitability
by low-frequency transcranial magnetic stimulation. Neurology, 48: 1398–1403
Chibbaro G, Daniele M, Alagona G, Di Pasquale C, Cannavo M, Rapisarda V, Bella R, Pennisi G. 2005. Repetitive transcranial
magnetic stimulation in schizophrenic patients reporting auditory hallucinations. Neurosci Lett, 383 (1–2): 54–57.
Cincotta M, Giovannelli F, Borgheresi A, Balestrieri F, Toscani L, Zaccara G, Carducci F, Viggiano MP, Rossi S. 2010. Optically
tracked neuronavigation increases the stability of hand-held focal coil positioning: evidence from transcranial magnetic
stimulation-induced electrical field measurements. Brain Stimul. 2010 Apr;3(2):119-23.
Cicero BA, BorodJC, SantschiC, ErhanHM, Obler LK, Agosti RM, Welkowitz J, Grunwald IS. 1999. Emotional versus
nonemotional lexical perception in patients with right and left brain damage. Neuropsychiatry Neuropsychol Behav Neurol,
12: 255-264.
Citron FM, Gray MA, Critchley HD, Weekes BS, Ferstl EC. 2014. Emotional valence and arousal affect reading in an
interactive way: Neuroimaging evidence for an approach-withdrawal framework. Neuropsychologia, 56: 79–89.
Citron FMM. 2012. Neural correlates of written emotion word processing: A review of recent electrophysiological and
hemodynamic neuroimaging studies. Brain and Language, 122(3): 211–226.
Coburger J, Musahl C, Henkes H, Horvath-Rizea D, Bittl M, Weissbach C, Hopf N. 2013. Comparison of navigated transcranial
magnetic stimulation and functional magnetic resonance imaging for preoperative mapping in rolandic tumor surgery.
Neurosurg Rev, 36(1): 65-75; discussion 75-6.
Cohen D, Cuffin B. 1991. Developing a more focal magnetic stimulator. J Clin Neurophysiol, 8(1): 102–111.
Cohen-Seat G, Gastaut H, Faure, J, Heuyer G. 1954. Etudes expérimentales de l'activité nerveuse pendant la projection
cinématographique. Rev Int Filmol. 5: 7–64.
Corbetta M, Shulman GL. 2002. Control of goal-directed and stimulus-driven attention in the brain. Nat Rev Neurosci, 3:
201–215.
Corcoran KM, Woody SR, Tolin DF. 2008. Recognition of facial expressions in obsessive-compulsive disorder. J Anxiety
Disord, 22(1): 56-66.
Corthout E, Barker AT, Cowey A. 2001. Transcranial magnetic stimulation: Which part of the current waveform causes the
stimulation? Exp Brain Res, 141: 128–132.
D
Damasio AR. 1996. The somatic marker hypothesis and the possible functions of the prefrontal cortex. Phil Trans R Soc Lond.
351: 1413–1420.
Darwin C. 1872. The Expression of the Emotions in Man and Animals. John Murray, London.
88
Das A, Anupa AV, Radhakrishnan A. 2013. Reduced plastic brain responses to repetitive transcranial magnetic stimulation in
severe obstructive sleep apnea syndrome. Sleep Medicine, Available online. http://dx.doi.org/10.1016/j.sleep.2013.04.008
Davidson RJ, Fox NA. 1982. Asymmetrical brain activity discriminates between positive and negative affective stimuli in
human infants. Science, 218 (4578): 1235–1237.
Davidson RJ. 1992. Anterior cerebral asymmetry and the nature of emotion. Brain Cogn, 20(1): 125–151.
Davidson RJ. 1998. Affective style and affective disorders: Perspectives from affective neuroscience. Cognition and Emotion,
12: 307–330.
Davidson RJ. 2000. Affective style, psychopathology, and resilience: brain mechanisms and plasticity. Am Psychol,
55(11):1196-214.
Davidson RJ. 2003. Affective neuroscience and psychophysiology: toward a synthesis. Psychophysiology, 40(5): 655-65.
Davidson RJ. 2004. What does the prefrontal cortex ‘‘do’’ in affect: Perspectives on frontal EEG asymmetry research.
Biological Psychology, 67: 219-233.
Day BL, Dressler D, Maertens de Noordhout A, Marsden CD, Nakashima K, Rothwell JC, Thompson PD. Electric and magnetic
stimulation of human motor cortex: surface EMG and single motor unit responses. J Physiol 1989; 412: 449–73.
Deblieck C, Thompson B, Iacoboni M, Wu AD. 2008. Correlation between motor and phosphene thresholds: a transcranial
magnetic stimulation study. Hum Brain Mapp. 2008 Jun;29(6): 662-70.
Dehaene S, Cohen L. 2011. The unique role of the visual word form area in reading. Trends Cogn Sci, 15(6): 254-62.
Deng Z-D, Lisanby SH, Peterchev A V. 2010. Transcranial magnetic stimulation in the presence of deep brain stimulation
implants: Induced electrode currents. Conf Proc IEEE Eng Med Biol Soc, 2010:6821–4.
Derejko M, Rakowicz M, Antczak J, Inglot E, Niewiadomska M. 2013. Corticomotor excitability in drug-naive patients with
Parkinson disease. Neurologia i Neurochirurgia Polska, 47(2):109-15.
Di Lazzaro V, Oliviero A, Mazzone P, Insola A, Pilato F, Saturno E, Accurso A, Tonali P, Rothwell JC. 2001. Comparison of
descending volleys evoked by monophasic and biphasic magnetic stimulation of the motor cortex in conscious humans. Exp
Brain Res, 141(1): 121-7.
Di Lazzaro V, Oliviero A, Pilato F, Saturno E, Dileone M, Mazzone P, Insola A, Tonali PA, Rothwell JC, 2004. The physiological
basis of transcranial motor cortex stimulation in conscious humans. Clin. Neurophysiol. 115, 255–266.
di Pellegrino G, Fadiga L, Fogassi L, Gallese V, Rizzolatti G. 1992. Understanding motor events: a neurophysiological study.
Exp Brain Res, 91(1):176-80.
Dillon DG, Pizzagalli DA. 2013. Evidence of successful modulation of brain activation and subjective experience during
reappraisal of negative emotion in unmedicated depression. Psychiatry Res, 212(2):99-107.
Dimberg U. 1982. Facial reactions to facial expressions. Psychophysiology, 19: 643–647.
Dolan RJ, Fletcher P, Morris J, Kapur N, Deakin JF, Frith CD. 1996. Neural activation during covert processing of positive
emotional facial expressions. Neuroimage, 4(3Pt1): 194-200.
Drager B, Breitenstein C, Helmke U, Kamping S, Knecht S. 2004. Specific and nonspecific effects of transcranial magnetic
stimulation on picture-word verification. Eur. J. Neurosci. 20: 1681–1687.
Dresler T, Meriau K, Heekeren HR, van der Meer E. 2009. Emotional Stroop task: effect of word arousal and subject anxiety
on emotional interference. Psychological Research, 73(3): 364–371.
E
89
Eche J, Mondino M, Haesebaert F, Saoud M, Poulet E, Brunelin J. 2012. Low- vs High-Frequency Repetitive Transcranial
Magnetic Stimulation as an Add-On Treatment for Refractory Depression. Front Psychiatry, 3:13.
Edwards J, Jackson HJ, Pattison PE. 2002. Emotion recognition via facial expression and affective prosody in schizophrenia: a
methodological review. Clin Psychol Rev 22: 789–832.
Edwardson M, Fetz EE, Avery DH. 2011. Seizure produced by 20 Hz transcranial magnetic stimulation during isometric
muscle contraction in a healthy subject. Clin Neurophysiol, 122(11): 2326-7.
Eimer M, Holmes A. 2002. An ERP study on the time course of emotional face processing. Neuroreport, 13(4): 427-31.
Ekman P, Friesen WV. 1971. "Constants across cultures in the face and emotion". Journal of Personality and Social
Psychology 17: 124–129.
Ekman P, Friesen WV. 1979. Pictures of Facial Affect. Palo Alto, CA. Consulting Psychologists .
Ekman P. 1973. Darwin and Facial Expression: a Century of Research in Review. New York (USA). Academic. P Ekman.
Epstein CM, Schwartzberg DG, Davey KR, Sudderth DB. (1990) Localizing the site of magnetic brain stimulation in humans.
Neurology; 40:666-670.
Epstein CM. Seizure or convulsive syncope during 1-Hz rTMS? Clin Neurophysiol 2006;117:2566–8.
Epstein CM, Schwartzberg DG, Davey KR, Sudderth D. 1990. Localizing the site of magnetic brain stimulation in humans.
Neurology, 40(4), 666–670.
Erhan H, Borod JC, Tenke CE, Bruder GE. 1998. Identification of emotion in a dichotic listening taskevent-related brain
potential and behavioral findings. Brain Cogn, 37 (2): 286–307.
F
Fadiga L, Fogassi L, Pavesi G, Rizzolatti G. 1995. Motor facilitation during action observation: a magnetic stimulation study. J
Neurophysiol, 73(6): 2608-11.
Faraday M. Experimental Researches in Electricity, vol.I Richard and John Edward Taylor, 1839; vol.II Richard and John
Edward Taylor, 1844; vol.III Richard Taylor and William Francis, 1855
Ferbert A, Priori A, Rothwell JC, Day BL, Colebatch JG, Marsden CD. 1992. Interhemispheric inhibition of the human motor
cortex. Journal of Physiology, 453: 525–546.
Fernández-Carriba S, Loeches A, Morcillo A, Hopkins WD. 2002. Asymmetry in facial expression of emotions by chimpanzees.
Neuropsychologia, 40(9): 1523-33.
Ferrier D. 1876. The Functions of the Brain. New York, NY, US: G P Putnam's Sons.
Forget P. 2011. Assessment of mean auditory hazard incurred by occupational exposure to impulse noise. European Annals
of Otorhinolaryngology, Head and Neck Diseases, 128(1): 14–17.
Fox MD, Halko MA, Eldaief MC, Pascual-Leone A. 2012. Measuring and manipulating brain connectivity with resting state
functional connectivity magnetic resonance imaging (fcMRI) and transcranial magnetic stimulation (TMS). NeuroImage,
62(4): 2232–2243.
Fried I, Mateer C, Ojemann G, Wohns R, Fedio P. 1982. Organization of visuospatial functions in human cortex. Evidence
from electrical stimulation. Brain, 105(Pt 2): 349-71.
Fritsch G, Hitzig E. Ueber die elektrische Erregbarkeit des Grosshirns. Archiv für Anatomie, Physiologie und
wissenschaftliche Medicin, 1870 : 300-332.
90
Frühholz S, Jellinghaus A, Herrmann M. 2011. Time course of implicit processing and explicit processing of emotional faces
and emotional words. Biol Psychol, 87(2): 265-74.
G
Gainotti G. 1969. Réaction catastrophiques et manifestations d’indifferénce au cours des atteintes cérébrales.
Neuropsychologia, 7: 195–204.
Gainotti G. 1972. Emotional behavior and hemispheric side of lesion. Cortex, 8: 41–55.
Gainotti G. 2012. Unconscious processing of emotions and the right hemisphere. Neuropsychologia, 50(2): 205-18.
Galvani L. 1791. De viribus electricitatis in motu musculari. Commentarius. Ex typographia Instituti Scientiarum, Bologna.
Gammaitoni L, Haenggi P, Jung P, Marchesoni F. 1998. Stochastic resonance. Rev. Mod. Phys. : 223–288.
Gastaut HJ, Bert J. 1954. EEG changes during cinematographic presentation. Electroencephalogr Clin Neurophysiol, 6: 433–
444.
George MS, Lisanby SH, Avery D, McDonald WM, Durkalski V, Pavlicova M, Anderson B, Nahas Z, Bulow P, Zarkowski P,
Holtzheimer PE 3rd, Schwartz T, Sackeim HA. 2010. Daily left prefrontal transcranial magnetic stimulation therapy for major
depressive disorder: a sham-controlled randomized trial. Arch Gen Psychiatry, 67(5):507-16.
George MS, Wassermann EM, Kimbrell TA, Little JT, Williams WE, Danielson AL, Greenberg BD, Hallett M, Post RM. 1997.
Mood improvement following daily left prefrontal repetitive transcranial magnetic stimulation in patients with depression:
A placebo-controlled crossover trial. Am J Psychol, 154: 1752–1756.
Gerwig M, Kastrup O, Meyer BU, Niehaus L. 2003. Evaluation of cortical excitability by motor and phosphene thresholds in
transcranial magnetic stimulation. Journal of the Neurological Sciences, 215(1–2): 75–78.
Geselowitz DB. 1998. The zero of potential. IEEE Engineering in Medicine and Biology Magazine, 17(1): 128–136.
Giovannelli F, Banfi C, Borgheresi A, Fiori E, Innocenti I, Rossi S, Zaccara G, Viggiano MP, Cincotta M. 2013. The effect of
music on corticospinal excitability is related to the perceived emotion: A transcranial magnetic stimulation study. Cortex,
49(3): 702–710.
Gough PM, Nobre AC, Devlin JT. 2005. Dissociating linguistic processes in the left inferior frontal cortex with Transcranial
Magnetic Stimulation. Journal of Neuroscience 25, 8010–8016.
Graves R, Landis T, Goodglass H. 1981. Laterality and sex differences for visual recognition of emotional and nonemotional
words. Neuropsychologia, 19(1): 95-102.
Gray JA, David AS. 1997. A specific neural substrate for perceiving facial expressions of disgust. Nature, 389(6650): 495-8.
Groppa S, Oliviero A, Eisen A, Quartarone A, Cohen LG, Mall V, Kaelin-Lang A, Mima T, Rossi S, Thickbroom GW, Rossini PM,
Ziemann U, Valls-Solé J, Siebner HR. 2012. A practical guide to diagnostic transcranial magnetic stimulation: report of an
IFCN committee. Clinical Neurophysiology : Official Journal of the International Federation of Clinical Neurophysiology,
123(5): 858–82.
H
Hamann S, Mao H. 2002. Positive and negative emotional verbal stimuli elicit activity in the left amygdala. Neuroreport, 13 :
15–19.
Hamann SB, Ely TD, Grafton ST, Kilts CD. 1999. Amygdala activity related to enhanced memory for pleasant and aversive
stimuli. Nat Neurosci, 2(3): 289-93.
Hansenne M, Laloyaux O, Mardaga S, Ansseau M. 2004. Impact of low frequency transcranial magnetic stimulation on
event-related brain potentials. Biol Psychol, 67(3):331-41.
91
Harlow JM. 1868. Recovery of the passage of an iron bar through the head. Publ Mass Med Soc, 2 : 327–334.
Hari R, Forss N, Avikainen S, Kirveskari E, Salenius S, Rizzolatti G. 1998. Activation of human primary motor cortex during
action observation: a neuromagnetic study. Proc Natl Acad Sci USA, 95(25): 15061-5.
Harris J, Clifford C, Miniussi C. The functional effect of transcranial magnetic stimulation: signal suppression or neural noise
generation? J Cogn Sci 2008;20:734–40.
Hatfield E, Cacioppo JT, Rapson RL. 1993. Emotional contagion. Curr Dir Psychol Sci, 2: 96–99.
Heller W, Nitschke JB, Lindsay DL. 1997. Neuropsychological correlates of arousal in self-reported emotion. Cognition and
Emotion, 11: 383–402.
Hennenlotter A, Schroeder U, Erhard P, Castrop F, Haslinger B, Stoecker D, Lange KW, Caballos-Baumann AO. 2005. A
common neural basis for receptive and expressive communication of pleasant facial affect. NeuroImage, 26:581–591.
Herbert C, Herbert BM, Pauli P. 2011. Emotional self-reference: brain structures involved in the processing of words
describing one's own emotions. Neuropsychologia, 49(10): 2947-56.
Hidding U, Bäumer T, Siebner HR, Demiralay C, Buhmann C, Weyh T, et al. 2006. MEP latency shift after implantation of
deep brain stimulation systems in the subthalamic nucleus in patients with advanced Parkinson’s disease. Mov Disord,
21:1471–6.
Hill AT, Fitzgibbon BM, Arnold SL, Rinehart NJ, Fitzgerald PB, Enticott PG. 2013. Modulation of putative mirror neuron
activity by both positively and negatively valenced affective stimuli: a TMS study. Behav Brain Res, 249:116-23.
Hoekert M, Bais L, Kahn RS, Aleman A. 2008. Time course of the involvement of the right anterior superior temporal gyrus
and the right fronto-parietal operculum in emotional prosody perception. PLoS One, 3(5):e2244.
Hoffman RE, Gueorguieva R, Hawkins KA, Varanko M, Boutros NN, Wu Y-T, Carroll K. 2005. Temporoparietal transcranial
magnetic stimulation for auditory hallucinations: safety, efficacy and moderators in a fifty patient sample. Biol Psychiatry,
58(2): 97–104.
Holler I, Siebner HR, Cunnington R, Gerschlager W. 2006. 5 Hz repetitive TMS increases anticipatory motor activity in the
human cortex. Neurosci Lett, 392(3):221-5.
Hsieh S, Hornberger M, Piguet O, Hodges JR. 2012. Brain correlates of musical and facial emotion recognition: evidence
from the dementias. Neuropsychologia, 50(8): 1814-22.
Huang YZ, Edwards MJ, Rounis E, Bhatia KP, Rothwell JC. 2005. Theta burst stimulation of the human motor cortex. Neuron,
45(2):201-6.
Hwang JH, Kim SH, Park CS, Bang SA, Kim SE. 2010. Acute high-frequency rTMS of the left dorsolateral prefrontal cortex and
attentional control in healthy young men. Brain Research, 1329: 152–158.
I
Iacoboni M, Woods RP, Brass M, Bekkering H, Mazziotta JC, Rizzolatti G. 1999. Cortical mechanisms of human imitation.
Science, 286(5449): 2526-8.
Iacoboni M, Dapretto M. 2006. The mirror neuron system and the consequences of its dysfunction. Nat Rev Neurosci,
7(12):942-51.
Ikeda T, Kurosawa M, Uchikawa C, Kitayama S, Nukina N. 2005. Modulation of monoamine transporter expression and
function by repetitive transcranial magnetic stimulation. Biochem Biophys Res Commun, 327(1): 218-24.
Ikeguchi M, Touge T, Kaji R, Deguchid K, Sasakid I, Tsukaguchid M, Takeuchie H,Kuriyamad S. 2005. Durable effect of very
low-frequency repetitive transcranial magnetic stimulation for modulating cortico-spinal neuron excitability. Int Congr
Series, 1278: 272–275
92
Ilmoniemi RJ, Virtanen J, Ruohonen J, Karhu J, Aronen HJ, Näätänen R, Katila T. (1997) Neuronal responses to magnetic
stimulation reveal cortical reactivity and connectivity. Neuroreport, 8(16): 3537–40.
Inase M, Tokuno H, Nambu A, et al. Origin of thalamocortical projections to the presupplementary motor area (pre-SMA) in
the macaque monkey. Neurosci Res 1996; 25: 217–227.
Isenberg N, Silbersweig D, Engelien A, Emmerich S, Malavade K, Beattie B, Leon AC, Stern E. 1999. Linguistic threat activates
the human amygdala. Proc Natl Acad Sci USA, 96(18): 10456-9.
Ishibashi R, Lambon Ralph MA, Saito S, Pobric G. 2011. Different roles of lateral anterior temporal lobe and inferior parietal
lobule in coding function and manipulation tool knowledge: Evidence from an rTMS study. Neuropsychologia, 49(5): 1128–
1135.
Ives JR, Rotenberg A, Poma R, Thut G, Pascual-Leone A. 2006. Electroencephalographic recording during transcranial
magnetic stimulation in humans and animals. Clin Neurophysiol 117: 1870–1875
J
Jelić MB, Stevanović VB, Milanović SD, Ljubisavljević MR, Filipović SR. 2013. Transcranial magnetic stimulation has no
placebo effect on motor learning. Clin Neurophysiol. 124(8):1646-51.
Jennum P, Friberg L, Fuglsang-Frederiksen A, Dam M. 1994. Speech localization using repetitive transcranial magnetic
stimulation. Neurology, 44(2): 269-73.
Johansson B. Measuring methods and risk levels for impulse noise in current standards and regulations. Scand Audiol Suppl
(Suppl 12) (1980), pp. 55–60
Jong Seok Bae, Neil G. Simon, Parvathi Menon, Steve Vucic, Matthew C. Kiernan. 2013. The Puzzling Case of
Hyperexcitability in Amyotrophic Lateral Sclerosis. J Clin Neurol., 9(2): 65-74.
K
Kähkönen S, Kesäniemi M, Nikouline VV, Karhu J, Ollikainen M, Holi M, Ilmoniemi RJ. 2001. Ethanol modulates cortical
activity: direct evidence with combined TMS and EEG. Neuroimage, 14: 322–328.
Kähkönen S, Wilenius J, Komssi S, Ilmoniemi RJ. 2004. Distinct differences in cortical reactivity of motor and prefrontal
cortices to magnetic stimulation. Clin. Neurophysiol., 115 : 583–588
Kammer T, Beck S, Thielscher A, Laubis-Herrmann U, Topka H. (2001). Motor thresholds in humans: a transcranial magnetic
stimulation study comparing different pulse waveforms, current directions and stimulator types. Clinical neurophysiology :
official journal of the International Federation of Clinical Neurophysiology, 112(2), 250–8.
Kano T, Kobayashi M, Ohira T, Yoshida K. 2012. Speech-induced modulation of interhemispheric inhibition. Neurosci Lett,
531(2):86-90.
Kanske P, Kotz SA. 2007. Concreteness in emotional words: ERP evidence from a hemifield study. Brain Res, 1148: 138-48.
Kanske P, Kotz SA. 2011. Emotion triggers executive attention: anterior cingulate cortex and amygdala responses to
emotional words in a conflict task. Human Brain Mapping, 32: 198–208.
Karlsson KA, Windischberger C, Gerstl F, Mayr W, Siegel JM, Moser E. 2010. Modulation of hypothalamus and amygdalar
activation levels with stimulus valence. Neuroimage, 51(1): 324-8.
Kawasaki H, Kaufman O, Damasio H, et al. Single-neuron responses to emotional visual stimuli recorded in human ventral
prefrontal cortex. Nat Neurosci 2001; 4: 15–16.
Kensinger EA, Anderson A, Growdon JH, Corkin S. 2004. Effects of Alzheimer disease on memory for verbal emotional
information. Neuropsychologia, 42(6): 791-800.
93
Kernell D, Chien-Ping W. Post-synaptic effects of cortical stimulation on forelimb motoneurones in the baboon. J Physiol
1967; 191: 673-90.
Keuper K, Zwanzger P, Nordt M, Eden A, Laeger I, Zwitserlood P, Kissler J, Junghöfer M, Dobel C. 2014. How 'love' and 'hate'
differ from 'sleep': Using combined electro/magnetoencephalographic data to reveal the sources of early cortical responses
to emotional words. Hum Brain Mapp, 35(3): 875-88.
Khedr EM, Kotb H, Kamel NF, Ahmed MA, Sadek R, Rothwell JC. 2005. Longlasting antalgic effects of daily sessions of
repetitive transcranial magnetic stimulation in central and peripheral neuropathic pain. J Neurol Neurosurg Psychiatry,
76(6): 833-8.
Kim H, Somerville LH, Johnstone T, Alexander AL, Whalen PJ. 2003. Inverse amygdala and medial prefrontal cortex
responses to surprised faces. Neuroreport, 14(18): 2317-22.
Kim H, Somerville LH, Johnstone T, Polis S, Alexander AL, Shin LM, Whalen PJ. 2004. Contextual modulation of amygdala
responsivity to surprised faces. J Cogn Neurosci, 16(10): 1730-45.
Kim YH, Min SJ, Ko MH, Park JW, Jang SH, Lee PK. 2005. Facilitating visuospatial attention for the contralateral hemifield by
repetitive TMS on the posterior parietal cortex. Neuroscience Letters, 382(3): 280–285.
Kissler J, Assadollahi R, Herbert C. 2006. Emotional and semantic networks in visual word processing: Insights from ERP
studies. Progress in Brain Research, 156: 147–183.
Kissler J, Koessler S. 2011. Emotionally positive stimuli facilitate lexical decisions-an ERP study. Biol Psychol, 86(3): 254-264.
Kober H, Barrett LF, Joseph J, Bliss-Moreau E, Lindquist K, Wager TD. 2008. Functional grouping and cortical-subcortical
interactions in emotion: a meta-analysis of neuroimaging studies. Neuroimage, 42(2): 998-1031.
Kozel FA, Nahas Z, deBrux C, Molloy M, Lorberbaum JP, Bohning D, Risch SC, George MS (2000) How coil-cortex distance
relates to age, motor threshold, and antidepressant response to repetitive transcranial magnetic stimulation. J
Neuropsychiatry Clin Neurosci 12:376–384.
Kratz O, Studer P, Barth W, Wangler S, Hoegl T, Heinrich H, Moll GH. 2011. Seizure in a nonpredisposed individual induced
by single-pulse transcranial magnetic stimulation. J ECT, 27(1): 48-50.
Krolak-Salmon P, Fischer C, Vighetto A, Mauguière F. 2001. Processing of facial emotional expression: spatiotemporal data
as assessed by scalp event-related potentials. Eur J Neurosci, 13: 987–994.
Krolak-Salmon P, Hénaff MA, Isnard J, Tallon-Baudry C, Guénot M, Vighetto A, Bertrand O, Mauguière F. 2003. An attention
modulated response to disgust in human ventral anterior insula. Ann Neurol, 53(4): 427-8.
Krolak-Salmon P, Henaff MA, Vighetto A, Bertrand O, Mauguière F. 2004. Early amygdala reaction to fear spreading in
occipital, temporal, and frontal cortex: a depth electrode ERP study in human. Neuron, 42(4): 665-76.
Krolak-Salmon P, Hénaff MA, Vighetto A, Bauchet F, Bertrand O, Mauguière F, Isnard J. 2006. Experiencing and Detecting
Hapiness in Humans: The Role of the Supplementary Motor Area. Ann Neurol, 59: 196-99.
Krummenacher P, Candia V, Folkers G, Schedlowski M, Schönbächler G. 2010. Prefrontal cortex modulates placebo
analgesia. Pain, 148(3):368-74.
Kühn AA, Brandt SA, Kupsch A, Trottenberg T, Brocke J, Irlbacher K, et al.2004. Comparison of motor effects following
subcortical electrical stimulation through electrodes in the globus pallidus internus and cortical transcranial magnetic
stimulation. Exp Brain Res, 155: 48–55.
Kühn AA, Trottenberg T, Kupsch A, Meyer BU. 2002. Pseudo-bilateral hand motor responses evoked by transcranial
magnetic stimulation in patients with deep brain stimulators. Clin Neurophysiol, 113:341–5.
Kumar R, Chen R, Ashby P. 1999. Safety of transcranial magnetic stimulation in patients with implanted deep brain
stimulators. Mov Disord, 14: 157–8.
94
L
Lalande S, Braun CM, Charlebois N, Whitaker HA. 1992. Effects of right and left hemisphere cerebrovascular lesions on
discrimination of prosodic and semantic aspects of affect in sentences. Brain Lang, 42(2): 165-86.
Landis T, Graves R, Goodglass H. 1982. Aphasic reading and writing: Possible evidence for right hemisphere participation.
Cortex, 18: 105-112.
Landis T, Regard M, Graves R, Goodglass H. 1983. Semantic paralexia: A release of right hemisphere function from left
hemispheric control? Neuropsychologia, 21: 359-364.
Lapicque, L. (1931) Has the muscular substance a longer chronaxie than the nervous substance? The Journal of Physiology
73: 189-214.
Lau HC, Rogers RD, Haggard P, Passingham RE. 2004. Attention to intention. Science, 303: 1208-1210.
Lawrence AD, Goerendt IK, Brooks DJ. 2007. Impaired recognition of facial expressions of anger in Parkinson's disease
patients acutely withdrawn from dopamine replacement therapy. Neuropsychologia, 45(1): 65-74.
LeDoux JE. 1993. Emotional memory systems in the brain. Behav Brain Res. 58: 69-79.
Lefaucheur JP, Hatem S, Nineb A, Ménard-Lefaucheur I, Wendling S, Keravel Y, Nguyen JP. 2006. Somatotopic organization
of the analgesic effects of motor cortex rTMS in neuropathic pain. Neurology, 67(11): 1998-2004.
Lefaucheur JP. 2006. The use of repetitive transcranial magnetic stimulation (rTMS) in chronic neuropathic pain.
Neurophysiol Clin. 2006 May-Jun;36(3):117-24.
Lefrou G. 1956. Contribution à l'étude de la mimique chez les anthropoïdes : le rire chez les chimpanzés. Bulletins et
Mémoires de la Société d'anthropologie de Paris, 7 lien Numéro 7-3-4 : 283-288.
Lehmann D, Michel CM. 2011. EEG-defined functional microstates as basic building blocks of mental processes. Clin
Neurophysiol, 122(6): 1073-4.
Lehmann D, Skrandies W. 1980. Reference-free identification of components of checkerboard-evoked multichannel
potential fields. Electroencephalography and Clinical Neurophysiology, 48(6): 609–621.
Lewis PA, Critchley HD, Rotshtein P, Dolan RJ. 2007. Neural correlates of processing valence and arousal in affective words.
Cereb Cortex, 17(3): 742-8.
Li CT, Su TP, Hsieh JC, Ho ST. 2013. Efficacy and practical issues of repetitive transcranial magnetic stimulation on chronic
medically unexplained symptoms of pain. Acta Anaesthesiol Taiwan, 51(2): 81-7.
Li X, Hartwell KJ, Owens M, Lematty T, Borckardt JJ, Hanlon CA, Brady KT, George MS. 2013. Repetitive transcranial
magnetic stimulation of the dorsolateral prefrontal cortex reduces nicotine cue craving. Biol Psychiatry, 73(8): 714-20.
Li X, Malcolm RJ, Huebner K, Hanlon CA, Taylor JJ, Brady KT, George MS, See RE. 2013. Low frequency repetitive transcranial
magnetic stimulation of the left dorsolateral prefrontal cortex transiently increases cue-induced craving for
methamphetamine: a preliminary study. Drug Alcohol Depend, 133(2): 641-6.
Lidzba K, Schwilling E, Grodd W, Krägeloh-Mann I, Wilke M. 2011. Language comprehension vs. language production: age
effects on fMRI activation. Brain Lang, 119(1): 6-15.
Likowski KU, Mühlberger A, Gerdes AB, Wieser MJ, Pauli P, Weyers P. 2012. Facial mimicry and the mirror neuron system:
simultaneous acquisition of facial electromyography and functional magnetic resonance imaging. Front Hum Neurosci,
6:214.
Lioumis P, Zhdanov A, Mäkelä N, Lehtinen H, Wilenius J, Neuvonen T, Hannula H, Deletis V, Picht T, Mäkelä JP. 2012. A novel
approach for documenting naming errors induced by navigated transcranial magnetic stimulation. J Neurosci Methods,
204(2): 349-54.
95
Lipton RB, Dodick DW, Silberstein SD, Saper JR, Aurora SK, Pearlman SH, Fischell RE, Ruppel PL, Goadsby PJ. 2010. Singlepulse transcranial magnetic stimulation for acute treatment of migraine with aura: a randomised, doubleblind, parallelgroup, sham-controlled trial. Lancet Neurol, 9(4):373-80.
Lipton RB, Pearlman SH. 2010. Transcranial magnetic simulation in the treatment of migraine. Neurotherapeutics, 7(2):20412.
Lisanby SH, Gutman D, Luber B, Schroeder C, Sackeim HA. 2001. Sham TMS: intracerebral measurement of the induced
electrical field and the induction of motor-evoked potentials. Biol Psychiatry. 1;49(5):460-3.
Liu W, Miller BL, Kramer JH, Rankin K, Wyss-Coray C, Gearhart R, Phengrasamy L, Weiner M, Rosen HJ. 2004. Behavioral
disorders in the frontal and temporal variants of frontotemporal dementia. Neurology, 62: 742–748.
Liu WH, Wang LZ, Zhao SH, Ning YP, Chan RC. 2012. Anhedonia and emotional word memory in patients with depression.
Psychiatry Res, 200(2-3): 361-7.
Lorenzano C, Dinapoli L, Gilio F, Suppa A, Bagnato S, Currà A, Inghilleri M, Berardelli A. 2006. Motor cortical excitability
studied with repetitive transcranial magnetic stimulation in patients with Huntington's disease. Clin Neurophysiol, 117(8):
1677-81.
Luber B, Lisanby SH. 2013. Enhancement of human cognitive performance using transcranial magnetic stimulation (TMS).
NeuroImage, Available online 13 June 2013.
Luppino G, Matelli M, Camarda R, Rizzolatti G. 1993. Corticocortical connections of area F3 (SMA-proper) and area F6 (preSMA) in the macaque monkey. J Comp Neurol, 338(1):114-40.
M
Maccabee PJ, Amassian VE, Eberle LP, Cracco RQ. Magnetic coil stimulation of straight and bent amphibian and mammalian
peripheral nerve in vitro: locus of excitation. J Physiol 1993;460:201–19.
Maclean PD. 1952. Some psychiatric implications of physiological studies on frontotemporal portion of limbic system
(visceral brain). Electroencephalogr Clin Neurophysiol. 4(4): 407-18.
Maeda F, Keenan JP, Tormos JM, Topka H, Pascual-Leone A. 2000. Modulation of corticospinal excitability by repetitive
transcranial magnetic stimulation. Clin. Neurophysiol., 111 : 800–805
Maizey L, Allen CP, Dervinis M, Verbruggen F, Varnava A, Kozlov M, Adams RC, Stokes M, Klemen J, Bungert A, Hounsell CA,
Chambers CD. 2013. Comparative incidence rates of mild adverse effects to transcranial magnetic stimulation. Clin
Neurophysiol, 124(3): 536-44.
Malenka RC, Bear MF. 2004. LTP and LTD: an embarrassment of riches. Neuron, 44: 5-21.
Mandal MK, Mohanty A, Pandey R, Mohanty S. 1996. Emotion-specific processing deficit in focal brain-damaged patients.
Int J Neurosci, 84(1–4): 87–95.
Marneweck M, Loftus A, Hammond G. 2013. Psychophysical measures of sensitivity to facial expression of emotion. Front
Psychol, 4: 63.
Martin L, Borckardt JJ, Reeves ST, Frohman H, Beam W, Nahas Z, Johnson K, Younger J, Madan A, Patterson D, George M.
2013. A pilot functional MRI study of the effects of prefrontal rTMS on pain perception. Pain Med, 14(7):999-1009.
Maskill D, Murphy K, Mier A, Owen M, Guz A. Motor cortical representa- tion of the diaphragm in man. J Physiol (Lond)
1991; 443:105–121.
Massimini M, Ferrarelli F, Huber R, Esser SK, Singh H, Tononi G. (2005) Breakdown of cortical effective connectivity during
sleep. Science (New York, N.Y.), 309(5744): 2228–32.
96
Mattavelli G, Cattaneo Z, Papagno C. 2011. Transcranial magnetic stimulation of medial prefrontal cortex modulates face
expressions processing in a priming task. Neuropsychologia. 49(5): 992-8.
Mattavelli G, Rosanova M, Casali AG, Papagno C, Romero Lauro LJ. 2013. Top-down interference and cortical
responsiveness in face processing: a TMS-EEG study. Neuroimage, 76:24-32.
Matteucci C. 1844. Traité des phénomènes électro-physiologiques des animaux, suivi d’Études Anatomiques sur le Système
nerveux et sur l’Organe Électrique de la torpille. Fortin et Masson, Paris.
Maxwell JC. 1861. On physical lines of force. Philosophical Magazine. The London, Edinburgh, and Dublin Philosophical
Magazine and Journal of Science, 21(4): 161-75; 281-91; 338-48. London. Taylor and Francis.
McConnell KA, Nahas Z, Shastri A, Lorberbaum JP, Kozel FA, Bohning DE, George MS (2001) The transcranial magnetic
stimulation motor threshold depends on the distance from coil to underlying cortex: a replication in healthy adults
comparing two methods of assessing the distance to cortex. Biol Psychiatry 49:454–549
Meyer BU, Röricht S, Gräfin von Einsiedel H, Kruggel F, Weindl A. 1995. Inhibitory and excitatory interhemispheric transfers
between motor cortical areas in normal humans and patients with abnormalities of the corpus callosum. Brain, 118(Pt2):
429-40.
Mhalla A, Baudic S, Ciampi de Andrade D, Gautron M, Perrot S, Teixeira MJ, Attal N, Bouhassira D. 2011. Long-term
maintenance of the analgesic effects of transcranial magnetic stimulation in fibromyalgia. Pain, 152(7): 1478-85.
Michel CM, Murray MM, Lantz G, Gonzalez S, Spinelli L, Grave de Peralta R. 2004. Review EEG source imaging. Clin
Neurophysiol, 115(10): 2195-222.
Michel CM, Murray MM. 2012. Towards the utilization of EEG as a brain imaging tool. NeuroImage, 61(2): 371-385.
Michelucci R, Valzania F, Passarelli D, Santangelo M, Rizzi R, Buzzi AM, Tempestini A, Tassinari CA. 1994. Rapid-rate
transcranial magnetic stimulation and hemispheric language dominance: usefulness and safety in epilepsy. Neurology,
44(9): 1697-700.
Miniussi C, Thut G. 2009. Combining TMS and EEG offers new prospects in cognitive neuroscience. Brain Topogr, 22(4):
249–56.
Misra UK, Kalita J, Bhoi SK. 2013. High-rate repetitive transcranial magnetic stimulation in migraine prophylaxis: a
randomized, placebo-controlled study. J Neurol. 2013 [Epub ahead of print]
Mohr C, Michel CM, Lantz G, Ortigue S, Viaud-Delmon I, Landis T. 2005. Brain state-dependent functional hemispheric
specialization in men but not in women. Cereb Cortex, 15(9): 1451-8.
Moore A, Gorodnitsky I, Pineda J. 2012. EEG mu component responses to viewing emotional faces. Behav Brain Res, 226(1):
309-16.
Mori F, Kusayanagi H, Monteleone F, Moscatelli A, Nicoletti CG, Bernardi G, Centonze D. 2013. Short interval intracortical
facilitation correlates with the degree of disability in multiple sclerosis. Brain Stimul. 6(1):67-71.
Moriguchi Y, Touroutoglou A, Dickerson BC, Barrett LF. 2013. Sex differences in the neural correlates of affective experience.
Soc Cogn Affect Neurosci, published online 2013 Apr 18.
Morris J S, Frith CD, Perrett DI, Rowland D, Young AW, Calder AJ, Dolan RJ. 1996. A differential neural response in the
human amygdala to fearful and happy facial expressions. Nature, 383(6603): 812-5.
Morris JS, Ohman A, Dolan RJ. 1998. Conscious and unconscious emotional learning in the human amygdala. Nature,
393(6684): 467-70.
Mostofsky SH, Simmonds DJ. 2008. Response inhibition and response selection: two sides of the same coin. J Cogn Neurosci,
20(5): 751-61.
97
Mottaghy FM, Pascual-Leone A, Kemna LJ, Töpper R, Herzog H, Müller-Gärtner HW, Krause BJ. 2003. Modulation of a brain–
behavior relationship in verbal working memory by rTMS. Cognitive Brain Research, 15(3): 241–249.
Muellbacher W, Ziemann U, Boroojerdi B, Hallett M. 2000. Effects of low-frequency transcranial magnetic stimulation on
motor excitability and basic motor behavior. Clinical Neurophysiology, 111(6): 1002–1007.
Mukamel R, Ekstrom AD, Kaplan J, Iacoboni M, Fried I. 2010. Single-neuron responses in humans during execution and
observation of actions. Curr Biol, 20(8):750-6.
Murray MM, Brunet D, Michel CM. 2008. Topographic ERP analyses: a step-by-step tutorial review. Brain Topogr, 20(4):
249-64.
N
Nachev P, Wydell H, O'neill K, Husain M, Kennard C. 2007. The role of the pre-supplementary motor area in the control of
action. Neuroimage, 36 Suppl 2:T155-63.
Nagarajan SS, Durand DM, Warman EN. Effects of induced electric fields on finite neuronal structures: a simulation study.
IEEE Trans Biomed Eng 1993;40:1175–88.
Nahas Z, Lomarev M, Roberts DR, Shastri A, Lorberbaum JP, Teneback C, McConnell K, Vincent DJ, Li X, George MS, Bohning
DE. 2001. Unilateral left prefrontal transcranial magnetic stimulation (TMS) produces intensity-dependent bilateral effects
as measured by interleaved BOLD fMRI. Biological Psychiatry, 50(9): 712–720.
Nardone R, Buffone E, Florio I, Tezzon F. 2005. Changes in motor cortex excitability during muscle fatigue in amyotrophic
lateral sclerosis. J Neurol Neurosurg Psychiatry, 76(3): 429–431.
Narumoto J, Okada T, Sadato N, Fukui K, Yonekura Y. 2001. Attention to emotion modulates fMRI activity in human right
superior temporal sulcus. Brain Res Cogn Brain Res, 12(2): 225-31.
Nauczyciel C, Hellier P, Morandi X, Blestel S, Drapier D, Ferre JC, Barillot C, Millet B. 2010. Assessment of standard coil
positioning in transcranial magnetic stimulation in depression. Psychiatry Res. 2011 Apr 30;186(2-3):232-8.
Netz J, Ziemann U, Homberg V. 1995. Hemispheric asymmetry of transcallosal inhibition in man. Exp Brain Res, 104: 527-33.
O
Ochsner KN, Gross JJ. 2005. The cognitive control of emotion. Trends Cogn Sci, 9(5): 242-249.
Oliveri M, Zhaoping L, Mangano GR, Turriziani P, Smirni D, Cipolotti L. 2010. Facilitation of bottom-up feature detection
following rTMS-interference of the right parietal cortex. Neuropsychologia, 48(4): 1003–1010.
Ortigue S, Michel CM, Murray MM, Mohr C, Carbonnel S, Landis T. 2004. Electrical neuroimaging reveals early generator
modulation to emotional words. Neuroimage, 21(4): 1242-51.
Ortu E, Ruge D, Deriu F, Rothwell JC. 2009. Theta Burst Stimulation over the human primary motor cortex modulates neural
processes involved in movement preparation. Clin Neurophysiol. 2009 Jun;120(6):1195-203..
P
Palazova M, Sommer W, Schacht A. 2013. Interplay of emotional valence and concreteness in word processing: an eventrelated potential study with verbs. Brain Lang, 125(3): 264-71.
Papez JW. 1937. A proposed mechanism of emotion. Arch Neurol Psychiatry. 38: 725–743.
Parks NA, Maclin EL, Low KA, Beck DM, Fabiani M, Gratton G. 2012. Examining cortical dynamics and connectivity with
simultaneous single-pulse transcranial magnetic stimulation and fast optical imaging. NeuroImage, 59(3): 2504–10.
Parr LA, Heintz M. 2009. Facial expression recognition in rhesus monkeys, Macaca mulatta. Anim Behav, 77(6): 1507-1513.
98
Pascual-Leone A, Davey NJ, Rothwell J, Wassermann EM, Puri BK, editors. Handbook of transcranial magnetic stimulation.
New York: Oxford University Press; 2002.
Pascual-Leone A, Gates JR, Dhuna A. 1991. Induction of speech arrest and counting errors with rapid-rate transcranial
magnetic stimulation. Neurology, 41(5): 697-702.
Pascual-Leone A, Houser CM, Reese K, Shotland LI, Grafman J, Sato S, Valls-Solé J, Brasil-Neto JP, Wassermann EM, Cohen
LG, et al. 1993. Safety of rapid-rate transcranial magnetic stimulation in normal volunteers. Electroencephalogr Clin
Neurophysiol, 89(2): 120-30.
Pascual-Leone A, Valls-Sollé J, Wassermann EM, Brasil-Neto J, Cohen LG, Hallett M. 1992. Effects of focal transcranial
magnetic stimulation on simple reaction time to acoustic, visual and somatosensory stimuli. Brain, 115(4): 1045-1059.
Pascual-Leone A, Walsh V. 2001. Fast backprojections from the motion to the primary visual area necessary for visual
awareness. Science, 292(5516): 510-2.
Pascual-Marqui RD, Lehmann D. Topographic maps, source localization inference, and the reference electrode: comments
on a paper by Desmedt et al. 1993. Electroencephalography and Clinical Neurophysiology, 88(6): 532–533.
Peelen MV, Atkinson AP, Vuilleumier P. 2010. Supramodal representations of perceived emotions in the human brain. J
Neurosci, 30(30): 10127-34.
Peng H, Zheng H, Li L, Liu J, Zhang Y, Shan B, Zhang L, Yin Y, Liu J, Li W, Zhou J, Li Z, Yang H, Zhang Z. 2012. Highfrequency
rTMS treatment increases white matter FA in the left middle frontal gyrus in young patients with treatment-resistant
depression. J Affect Disord, 136(3): 249-57.
Pérocheau D, Laroche F, Perrot S. 2013. Relieving pain in rheumatology patients: Repetitive transcranial magnetic
stimulation (rTMS), a developing approach. Joint Bone Spine, In press. pii: S1297-319X(13)00116-4.
Perria L, Rosadini G, Rossi GF. 1961. Determination of side of cerebral dominance with Amobarbital. Archives of Neurology,
4: 173–181.
Perry RJ, Rosen HR, Kramer JH, Beer JS, Levenson RL, Miller BL. 2001. Hemispheric dominance for emotions, empathy and
social behaviour: evidence from right and left handers with frontotemporal dementia. Neurocase, 7(2): 145-60.
Peterchev AV, Jalinous R, Lisanby SH. (2008). A transcranial magnetic stimulator inducing near-rectangular pulses with
controllable pulse width (cTMS). IEEE Trans Biomed Eng.; 55(1):257-66.
Peterchev AV, Murphy DL, Lisanby SH. 2011. Repetitive transcranial magnetic stimulator with controllable pulse parameters.
Journal of neural engineering, 8(3) : 036016.
Phillips ML, Young AW, Scott SK, Calder AJ, Andrew C, Giampietro V, Williams SC, Bullmore ET, Brammer M, Gray JA. 1998.
Neural responses to facial and vocal expressions of fear and disgust. Proc Biol Sci, 265(1408): 1809-17.
Phillips ML, Young AW, Senior C, Brammer M, Andrew C, Calder AJ, Bullmore ET, Perrett DI, Rowland D, Williams SC, Picht T,
Schmidt S, Brandt S, Frey D, Hannula H, Neuvonen T, Karhu J, Vajkoczy P, Suess O. 2011. Preoperative functional mapping
for rolandic brain tumor surgery: comparison of navigated transcranial magnetic stimulation to direct cortical stimulation.
Neurosurgery, 69(3): 581-8.
Ponari M, Conson M, D'Amico NP, Grossi D, Trojano L. 2012. Mapping correspondence between facial mimicry and emotion
recognition in healthy subjects. Emotion, 12(6): 1398-403.
Ponz A, Montant M, Liegeois-Chauvel C, Silva C, Braun M, Jacobs AM, Ziegler JC. 2013. Emotion processing in words: a test
of the neural re-use hypothesis using surface and intracranial EEG. Soc Cogn Affect Neurosci, In press.
Posamentier MT, Abdi H. 2003. Processing Faces and Facial Expressions. Neuropsychology Review, 13(3): 113-143.
99
Posner J, Russell JA, Gerber A, Gorman D, Colibazzi T, Yu S, Wang Z, Kangarlu A, Zhu H, Peterson BS. 2009. The
neurophysiological bases of emotion: An fMRI study of the affective circumplex using emotion-denoting words. Hum Brain
Mapp, 30(3):883-95.
Poulet E, Brunelin J, Bediou B, Bation R, Forgeard L, Dalery J, d'Amato T, Saoud M. 2005. Slow transcranial magnetic
stimulation can rapidly reduce resistant auditory hallucinations in schizophrenia. Biol Psychiatry, 57(2): 188–191.
Pourtois G, Grandjean D, Sander D, Vuilleumier P. 2004. Electrophysiological correlates of rapid spatial orienting towards
fearful faces. Cereb Cortex, 14(6): 619-33.
Pourtois G, Sander D, Andres M, Grandjean D, Reveret L, Olivier E, Vuilleumier P. 2004. Dissociable roles of the human
somatosensory and superior temporal cortices for processing social face signals. Eur J Neurosci, 20(12): 3507-15.
R
Ragazzoni A, Pirulli C, Veniero D, Feurra M, Cincotta M, Giovannelli F, Chiaramonti R, Lino M, Rossi S, Miniussi C. 2013.
Vegetative versus minimally conscious states: a study using TMS-EEG, sensory and event-related potentials. PloS one, 8(2),
e57069.
Recio G, Schacht A, Sommer W. 2014. Recognizing dynamic facial expressions of emotion: Specificity and intensity effects in
event-related brain potentials. Biol Psychol, 96: 111-25.
Reid P, Pridmore S. 2001. Improvement in chronic pain with transcranial magnetic stimulation. Aust N Z J Psychiatry, 35(2):
252.
Rellecke J, Palazova M, Sommer W, Schacht A. 2011. On the automaticity of emotion processing in words and faces: eventrelated brain potentials evidence from a superficial task. Brain Cogn, 77(1): 23-32.
Reuter-Lorenz P, Davidson RJ. 1981. Differential contributions of the two cerebral hemispheres to the perception of happy
and sad faces. Neuropsychologia, 19(4): 609-13.
Reuterskiöld C. 1991. The effects of emotionality on auditory comprehension in aphasia. Cortex, 27(4): 595-604.
Rives Bogart K, Matsumoto D. 2010. Facial mimicry is not necessary to recognize emotion: Facial expression recognition by
people with Moebius syndrome. Soc Neurosci, 5(2): 241-51.
Rizzolatti G, Fogassi L, Gallese V. 2001. Neurophysiological mechanisms underlying the understanding and imitation of
action. Nat Rev Neurosci, 2(9): 661-70.
Robotham L, Sauter DA, Bachoud-Lévi AC, Trinkler I. 2011. The impairment of emotion recognition in Huntington's disease
extends to positive emotions. Cortex, 47(7): 880-4.
Rochas V, Gelmini L, Krolak-Salmon P, Poulet E, Saoud M, Brunelin J, Bediou B. 2013. Disrupting pre-SMA activity impairs
facial happiness recognition: an event-related TMS study. Cereb Cortex, 23(7): 1517-25.
Rochas V, Rihs TA, Rosenberg N, Landis T, Michel CM. 2014. Very early processing of emotional words revealed in
temporoparietal junctions of both hemispheres by EEG and TMS. Exp Brain Res, 232(4): 1267–81.
Rogasch NC, Fitzgerald PB. 2013. Assessing cortical network properties using TMS-EEG. Human Brain Mapping, 34(7): 1652–
69.
Rolls ET. 2004. Convergence of sensory systems in the orbitofrontal cortex in primates and brain design for emotion. Anat
Rec A Discov Mol Cell Evol Biol. 281(1): 1212-25.
Romei V, Brodbeck V, Michel C, Amedi A, Pascual-Leone A, Thut G. 2008. Spontaneous fluctuations in posterior alpha-band
EEG activity reflect variability in excitability of human visual areas. Cereb Cortex, 18(9):2010-8.
100
Rösler J, Niraula B, Strack V, Zdunczyk A, Schilt S, Savolainen P, Lioumis P, Mäkelä J, Vajkoczy P, Frey D, Picht T. 2013.
Language mapping in healthy volunteers and brain tumor patients with a novel navigated TMS system: Evidence of tumorinduced plasticity. Clin Neurophysiol, Online 2013 Sep 16.
Rossi S, Ferro M, Cincotta M, Ulivelli M, Bartalini S, Miniussi C, Giovannelli F, Passero S. 2007. A real electromagnetic
placebo (REMP) device for sham transcranial magnetic stimulation (TMS). Clin Neurophysiol. 118(3):709-16.
Rossi S, Hallett M, Rossini PM, Pascual-Leone A. 2011. Screening questionnaire before TMS: an update. Clin Neurophysiol,
122(8): 1686.
Rossi S, Pasqualetti P, Rossini PM, Feige B, Ulivelli M, Glocker FX, Battistini N, Lucking CH, Kristeva-Feige R. 2000. Effects of
repetitive transcranial magnetic stimulation on movement-related cortical activity in humans. Cereb Cortex, 10(8):802-8.
Rossini PM, Barker AT, Berardelli A, Caramia MD, Caruso G, Cracco RQ, et al. Non- invasive electrical and magnetic
stimulation of the brain, spinal cord and roots: basic principles and procedures for routine clinical application. Report of an
IFCN committee. Electroencephalogr Clin Neurophysiol 1994; 91:79–92.
Rothkegel H, Sommer M, Paulus W. 2010a. Breaks during 5 Hz rTMS are essential for facilitatory after effects. Clin.
Neurophysiol., 121: 426–430.
Rothkegel H, Sommer M, Paulus W, Lang N. 2010b. Impact of pulse duration in single pulse TMS. Clinical neurophysiology :
official journal of the International Federation of Clinical Neurophysiology, 121(11), 1915–21.
Rothwell J. 2007. Transcranial magnetic stimulation as a method for investigating the plasticity of the brain in Parkinson's
Disease and dystonia. Parkinsonism & Related Disorders, 13(supp3): S417–S420.
Rothwell JC, Hallett M, Berardelli A, Eisen A, Rossini P, Paulus W. 1999. Magnetic stimulation: motor evoked potentials. The
International Federation of Clinical Neurophysiology. Electroencephalogr Clin Neurophysiol Suppl, 52:97-103.
Rottenberg J. 2005. Mood and emotion in major depression. Current Directions in Psychological Science, 14: 167–170.
Rottenberg J, Gross JJ, Gotlib IH. 2005. Emotion context insensitivity in major depressive disorder. Journal of Abnormal
Psychology, 114 : 627–639.
Ruohonen J, Ilmoniemi RJ. Physical principles for transcranial magnetic stimulation. In: Pascual-Leone A, Davey NJ, Rothwell
J, Wassermann EM, Puri BK, editors. Handbook of transcranial magnetic stimulation. New York: Oxford University Press;
2002.
Rushworth MF, Buckley MJ, Behrens TE, Walton ME, Bannerman DM. 2007. Functional organization of the medial frontal
cortex. Curr Opin Neurobiol, 17(2): 220-7.
S
Sack AT, Cohen Kadosh R, Schuhmann T, Moerel M, Walsh V, Goebel R. 2009. Optimizing functional accuracy of TMS in
cognitive studies: a comparison of methods. J Cogn Neurosci, 21(2): 207-21.
Sackeim HA, Gur RC, Saucy MC. 1978. Emotions are expressed more intensely on the left side of the face. Science,
202(4366): 434-6.
Saitoh Y, Yoshimine T. 2007. Stimulation of primary motor cortex for intractable deafferentation pain. Acta Neurochir Suppl,
97(Pt 2): 51-6.
Sampson SM, Rome JD, Rummans TA. Slow-frequency rTMS reduces fibromyalgia pain. Pain Med, 7(2): 115-8.
Sanjuan J, Lull JJ, Aguilar EJ, Martí-Bonmatí L, Moratal D, Gonzalez JC, Robles M, Keshavan MS. 2007. Emotional words
induce enhanced brain activity in schizophrenic patients with auditory hallucinations. Psychiatry Res, 154(1): 21-9.
Sato W, Kubota Y, Okada T, Murai T, Yoshikawa S, Sengoku A. 2002. Seeing happy emotion in fearful and angry faces:
qualitative analysis of facial expression recognition in a bilateral amygdala-damaged patient. Cortex, 38(5): 727-42.
101
Sato W, Kochiyama T, Yoshikawa S, Naito E, Matsumura M. 2004. Enhanced neural activity in response to dynamic facial
expressions of emotion: an fMRI study. Brain Res Cogn Brain Res, 20(1): 81-91.
Sato W, Yoshikawa S. 2007. Spontaneous facial mimicry in response to dynamic facial expressions. Cognition, 104(1): 1-18.
Sauseng P, Klimesch W, Gerloff C, Hummel FC. 2009. Spontaneous locally restricted EEG alpha activity determines cortical
excitability in the motor cortex. Neuropsychologia, 47(1):284-8.
Schacht A, Sommer W. 2009. Time course and task dependence of emotion effects in word processing. Cogn Affect
BehavNeurosci, 9(1):28-43.
Schambra HM, Sawaki L, Cohen LG. 2003. Modulation of excitability of human motor cortex (M1) by 1 Hz transcranial
magnetic stimulation of the contralateral M1. Clinical Neurophysiology, 114(1): 130–133.
Schrader LM, Stern JM, Fields TA, Nuwer MR, Wilson CL. 2005. A lack of effect from transcranial magnetic stimulation (TMS)
on the vagus nerve stimulator (VNS). Clin Neurophysiol, 116: 2501–4.
Schulz H, Ubelacker T, Keil J, Müller N, Weisz N. 2013 Feb 8 [Epub ahead of print]. Now I am Ready--Now I am not: The
Influence of Pre-TMS Oscillations and Corticomuscular Coherence on Motor-Evoked Potentials. Cereb Cortex.
Schürmann M, Nikouline VV, Soljanlahti S, Ollikainen M, Basar E, Ilmoniemi RJ. 2001. EEG responses to combined
somatosensory and transcranial magnetic stimulation. Clin Neurophysiol, 112: 19–24.
Schutter DJLG, van Honk J. 2003. Reductions in CI amplitude after repetitive transcranial magnetic stimulation (rTMS) over
the striate cortex. Cognitive Brain Research, 16(3): 488–491.
Schutter DJ, van Honk J. 2006. An electrophysiological link between the cerebellum, cognition and emotion: frontal theta
EEG activity to single-pulse cerebellar TMS. Neuroimage, 33(4): 1227-31.
Scott GG, O'Donnell PJ, Sereno SC. 2013. Emotion words and categories: evidence from lexical decision. Cogn Process,
online publication 2013 Nov 21.
Seitz RJ, Schäfer R, Scherfeld D, Friederichs S, Popp K, Wittsack HJ, Azari NP, Franz M. 2008. Valuating other people's
emotional face expression: a combined functional magnetic resonance imaging and electroencephalography study.
Neuroscience, 152(3): 713-22.
Sel A, Forster B, Calvo-Merino B. 2014. The Emotional Homunculus: ERP Evidence for Independent Somatosensory
Responses during Facial Emotional Processing. J Neurosci, 34(9):3263-7.
Seyal M, Ro T, Rafal R. 1995. Increased sensitivity to ipsilateral cutaneous stimuli following transcranial magnetic
stimulation of the parietal lobe. Ann Neurol, 38(2): 264-7.
Shimojima Y, Morita H, Nishikawa N, Kodaira M, Hashimoto T, Ikeda S. 2010. The safety of transcranial magnetic stimulation
with deep brain stimulation instruments. Parkinsonism & Related Disorders. 16(2): 127–131.
Short E, Borckardt JJ, George M, Beam W, Reeves ST. 2009. Non-invasive brain stimulation approaches to fibromyalgia pain.
J Pain Manag, 2(3):259–276.
Skrandies W. 2013. Electrophysiological correlates of connotative meaning in healthy children. Brain Topogr, Epub 2013 Aug
22.
Slotema CW, Aleman A, Daskalakis ZJ, Sommer IE. 2012. Meta-analysis of repetitive transcranial magnetic stimulation in the
treatment of auditory verbal hallucinations: update and effects after one month. Schizophr Res, 142(1-3): 40-5.
Slotema CW, Blom JD, de Weijer AD, Diederen KM, Goekoop R, Looijestijn J, Daalman K, Rijkaart AM, Kahn RS, Hoek HW,
Sommer IE. 2011. Can low-frequency repetitive transcranial magnetic stimulation really relieve medicationresistant
auditory verbal hallucinations? Negative results from a large randomized controlled trial. Biol Psychiatry, 69(5): 450-6.
102
Smith SD, Bulman-Fleming MB. 2005. An examination of the right-hemisphere hypothesis of the lateralization of emotion.
Brain Cogn, 57(2): 210-3.
Sommer M, Alfaro A, Rummel M, Speck S, Lang N, Tings T, Paulus W. 2006. Half sine , monophasic and biphasic transcranial
magnetic stimulation of the human motor cortex. Clinical Neurophysiology, 117 : 838–844.
Sparing R, Buelte D, Meister IG, Paus T, Fink GR. 2008. Transcranial magnetic stimulation and the challenge of coil
placement: a comparison of conventional and stereotaxic neuronavigational strategies. Hum Brain Mapp, 29(1): 82-96.
Sparing R, Hesse MD, Fink GR. 2010. Neuronavigation for transcranial magnetic stimulation (TMS): where we are and where
we are going. Cortex, 46(1):118-20.
Speer AM, Benson BE, Kimbrell TK, Wassermann EM, Willis MW, Herscovitch P, Post RM. 2009. Opposite effects of high and
low frequency rTMS on mood in depressed patients: relationship to baseline cerebral activity on PET. J Affect Disord,
115(3):386-94.
Speer AM, Kimbrell TA, Wassermann EM, D Repella J, Willis MW, Herscovitch P, Post RM. 2000. Opposite effects of high and
low frequency rTMS on regional brain activity in depressed patients. Biol Psychiatry, 48(12):1133-41.
Speer AM, Wassermann EM, Benson BE, Herscovitch P, Post RM. 2013. Antidepressant Efficacy of High and Low Frequency
rTMS at 110% of Motor Threshold versus Sham Stimulation over Left Prefrontal Cortex. Brain Stimul, online 2013 Jul 29.
Spoletini I, Marra C, Di Iulio F, Gianni W, Sancesario G, Giubilei F, Trequattrini A, Bria P, Caltagirone C, Spalletta G. 2008.
Facial emotion recognition deficit in amnestic mild cognitive impairment and Alzheimer disease. Am J Geriatr
Psychiatry,16(5): 389-98.
Sprengelmeyer R, Young AW, Pundt I, Sprengelmeyer A, Calder AJ, Berrios G, Winkel R, Vollmöeller W, Kuhn W, Sartory G,
Przuntek H. 1997. Disgust implicated in obsessive-compulsive disorder. Proceedings of the Royal Society of London, 264:
1767-1773.
Starck J, Toppila E, Pyykkö I. 2003. Impulse noise and risk criteria. Noise Health. 5, pp. 63–73.
Straube T, Sauer A, Miltner WHR. 2011. Brain activation during direct and indirect processing of positive and negative words.
Behavioural Brain Research, 222 : 66–72.
Strauss E. 1983. Perception of emotional words. Neuropsychologia, 21(1): 99-103.
T
Tarapore PE, Findlay AM, Honma SM, Mizuiri D, Houde JF, Berger MS, Nagarajan SS. 2013. Language mapping with
navigated repetitive TMS: proof of technique and validation. Neuroimage, 82:260-72.
Taylor PC, Nobre AC, Rushworth MF. 2007a. FEF TMS affects visual cortical activity. Cereb Cortex, 17:391–399.
Taylor PC, Nobre AC, Rushworth MF. 2007b. Subsecond changes in top down control exerted by human medial frontal
cortex during conflict and action selection: a combined transcranial magnetic stimulation electroencephalography study. J
Neurosci, 27:11343–11353.
Terao Y, Ugawa Y, Hanajima R, Machii K, Furubayashi T, Mochizuki H, Enomoto H, Shiio Y, Uesugi H, Iwata NK, Kanazawa I.
Predominant activation of I1-waves from the leg motor area by transcranial magnetic stimulation. Brain Res 2000;859:137–
146.
Terao Y, Ugawa Y, Sakai K, Uesaka Y, Kohara N, Kanazawa I. 301 Transcranial stimulation of the leg area of the motor cortex
in humans. Acta Neurol Scand 1994;89:378–383.
Terzian H, Cecotto S. 1959. Su un nuovo metodo per la determinazione e lo studio della dominanza emisferica. Giornale di
Psichiatria e Neuropatologia, 87: 889–924.
103
Thickbroom GW. Transcranial magnetic stimulation and synaptic plasticity: experimental framework and human models.
Experimental Brain Research, 180: 583–593, 2007.
Thut G, Northoff G, Ives JR, Kamitani Y, Pfennig A, Kampmann F, Schomer DL, Pascual-Leone A. 2003. Effects of single-pulse
transcranial magnetic stimulation (TMS) on functional brain activity: a combined event-related TMS and evoked potential
study. Clin Neurophysiol 114: 2071–2080.
Thut G, Veniero D, Romei V, Miniussi C, Schyns P, Gross J. 2011a. Rhythmic TMS causes local entrainment of natural
oscillatory signatures. Current Biology, 21(14): 1176–85.
Thut G, Schyns PG, Gross J. 2011b. Entrainment of perceptually relevant brain oscillations by non-invasive rhythmic
stimulation of the human brain. Front Psychol, 2: 170.
Trautmann SA, Fehr T, Herrmann M. 2009. Emotions in motion: dynamic compared to static facial expressions of disgust
and happiness reveal more widespread emotion-specific activations. Brain Res, 1284: 100-15.
Tringali S, Perrot X, Collet L, Moulin A. 2012. Repetitive transcranial magnetic stimulation: Hearing safety considerations.
Brain Stimulation, 5(3): 354–363.
Tringali S, Perrot X, Collet L, Moulin A. 2013. Related noise exposure and auditory consequence during transcranial
magnetic stimulation: new insights and review of the literature [Article in French]. Neurophysiol Clin. 2013 Jan;43(1):19-33.
Tucker DM, Stenslie CE, Roth RS, Shearer SL. 1981. Right frontal lobe activation and right hemisphere performance.
Decrement during a depressed mood. Arch Gen Psychiatr, 38(2): 169–174.
V
Vacherot F, Attarian S, Vaugoyeau M, Azulay JP. 2010. A motor cortex excitability and gait analysis on Parkinsonian patients.
Mov Disord, 25(16):2747-55.
van der Gaag C, Minderaa RB, Keysers C. 2007. Facial expressions: what the mirror neuron system can and cannot tell us.
Soc Neurosci, 2(3-4): 179-222.
van Honk J, Hermans EJ, d'Alfonso AA, Schutter DJ, van Doornen L, de Haan EH. 2002. A left-prefrontal lateralized,
sympathetic mechanism directs attention towards social threat in humans: evidence from repetitive transcranial magnetic
stimulation. Neurosci Lett, 319(2): 99-102.
Vaughan HG. 1982. The neural origins of human event-related potentials. Ann N Y Acad Sci, 388: 125-38.
Vergnes M, Karli P. 1970. Déclenchement d'un comportement d'agression par stimulation électrique de l'hypothalamus
médian chez le rat. Physiology & Behavior, 5(12): 1427–1430.
Vernet M, Walker L, Yoo WK, Pascual-Leone A, Chang BS. 2012. EEG onset of a seizure during TMS from a focus
independent of the stimulation site. Clin Neurophysiol, 123(10): 2106-8.
Viggiano MP, Giovannelli F, Borgheresi A, Feurra M, Berardi N, Pizzorusso T, Zaccara G, Cincotta M. 2008. Disruption of the
prefrontal cortex function by rTMS produces a category-specific enhancement of the reaction times during visual object
identification. Neuropsychologia, 46(11) : 2725–2731.
Villarreal M, Fridman EA, Amengual A, Falasco G, Gerschcovich ER, Ulloa ER, Leiguarda RC. 2008. The neural substrate of
gesture recognition. Neuropsychologia, 46(9): 2371-82.
Vinson D, Ponari M, Vigliocco G. 2013. How does emotional content affect lexical processing? Cogn Emot, online publication
2013 Nov 12.
Virtanen J, Ruohonen J, Näätänen R, Ilmoniemi RJ. 1999. Instrumentation for the easurement of electric brain responses to
transcranial magnetic stimulation. Med Biol Eng Comput 37:322–326
104
Vrana SR, Gross D. 2004. Reactions to facial expressions: effects of social context and speech anxiety on responses to
neutral, anger, and joy expressions. Biol Psychol, 66(1): 63-78.
Vytal K, Hamann S. 2010. Neuroimaging support for discrete neural correlates of basic emotions: a voxel-based metaanalysis. J Cogn Neurosci, 22(12): 2864-85.
W
Wager TD, Phan KL, Liberzon I, Taylor SF. 2003. Valence, gender, and lateralization of functional brain anatomy in emotion:
a meta-analysis of findings from neuroimaging. Neuroimage, 19(3): 513-31.
Wagner T, Gangitano M, Romero R, Théoret H, Kobayashi M, Anschel D, Ives J, et al. 2004. Intracranial measurement of
current densities induced by transcranial magnetic stimulation in the human brain. Neuroscience Letters, 354(2): 91–94.
Wang Y, Isoda M, Matsuzaka Y, Shima K, Tanji J. 2005. Prefrontal cortical cells projecting to the supplementary eye field and
presupplementary motor area in the monkey. Neurosci Res, 53(1): 1-7.
Wassermann EM, Wang B, Zeffiro TA, Sadato N, Pascual-Leone A, Toro C, Hallett M. 1996. Locating the Motor Cortex on the
MRI with Transcranial Magnetic Stimulation and PET. NeuroImage, 9(1): 1–9.
Wassermann EM. 1998. Risk and safety of repetitive transcranial magnetic stimulation: report and suggested guidelines
from the international workshop on the safety of repetitive transcranial magnetic stimulation. Electroencephalogr Clin
Neurophysiol, 108:1-16.
Weiduschat N, Habedank B, Lampe B, Poggenborg J, Schuster A, Haupt WF, Heiss WD, Thiel A. 2009. Localizing Broca's area
for transcranial magnetic stimulation: Comparison of surface distance measurements and stereotaxic positioning. Brain
Stimul, 2(2): 93-102.
Weiss C, Nettekoven C, Rehme AK, Neuschmelting V, Eisenbeis A, Goldbrunner R, Grefkes C. 2012. Mapping the hand, foot
and face representations in the primary motor cortex - Retest reliability of neuronavigated TMS versus functional MRI.
Neuroimage, 66C: 531-542.
Werner-Seidler A, Moulds ML. 2012. Mood repair and processing mode in depression. Emotion, 12(3): 470-8.
White AT, Vanhaitsma TA, Vener J, Davis SL. 2013. Effect of passive whole body heating on central conduction and cortical
excitability in multiple sclerosis patients and healthy controls. J Appl Physiol. 114(12):1697-704.
Williams JH, Whiten A, Suddendorf T, Perrett DI. 2001. Imitation, mirror neurons and autism. Neurosci Biobehav Rev, 25(4):
287-95.
Willems RM, Clevis K, Hagoort P. 2011. Add a picture for suspense: neural correlates of the interaction between language
and visual information in the perception of fear. Soc Cogn Affect Neurosci. 6(4): 404-16.
Willems RM, Labruna L, D'Esposito M, Ivry R, Casasanto D. 2011. A functional role for the motor system in language
understanding: evidence from theta-burst transcranial magnetic stimulation. Psychol Sci, 22(7):849-54.
Williams JM, Mathews A, MacLeod C. 1996. The emotional Stroop task and psychopathology. Psychological Bulletin, 120(1):
3–24.
Wilson-Mendenhall CD, Feldman Barrett L, Barsalou LW. 2013. Neural Evidence That Human Emotions Share Core Affective
Properties. Psychological Science, 24(6): 947-956.
Workman L, Peters S, Taylor S. 2000. Lateralisation of perceptual processing of pro- and anti-social emotions displayed in
chimeric faces. Laterality, 5(3): 237-49.
Z
Zangen A, Roth Y, Voller B, Hallett M. 2005. TMS of deep brain regions: evidence for efficacy of the H-coil. Clinical
Neurophysiology, 116: 775–779.
105
Zanto TP, Chadick JZ, Gazzaley A. 2014. Anticipatory alpha phase influences visual working memory performance.
Neuroimage, 85(2): 794-802.
Ziemann U. 2004. TMS induced plasticity in human cortex. Rev. Neurosci., 15: 253–266.
106
X. ARTICLES
These are the two articles submitted and accepted respectively in Oxford Journals Cerebral
Cortex 2013 volume 23 issue 7 pp 1517-1525 (Rochas et al., 2013 doi: 10.1093/cercor/bhs133) and in
Springer Experimental Brain Research 2014 volume 232 issue 4 pp 1267-1281 (Rochas et al., 2014
doi: 10.1007/s00221-014-3843-y).
1. The first study using TMS: on facial expression recognition
Cerebral Cortex July 2013;23:1517-1525 doi:10.1093/cercor/bhs133
Advance Access publication June 1, 2012
Disrupting Pre-SMA Activity Impairs Facial Happiness Recognition: An Event-Related TMS
Study
Vincent Rochas 1,2,, Lauriane Gelmini 1, Pierre Krolak-Salmon 3,4, Emmanuel Poulet 1, Mohamed Saoud 1, Jerome Brunelin 1,5 and
6,7
Benoit Bediou
1
EA4615-SIPAD, Université Lyon 1, CH Le Vinatier, Bron F-69677, France 2Functional Brain Mapping Laboratory, Department of
3
Fundamental Neuroscience, University of Geneva, CH-1206 Geneva, Switzerland Memory Center of Lyon, Hôpital des
Charpennes, Hospices Civils de Lyon, Lyon, France 4Lyon Neuroscience Research Center, INSERM U1028, CNRS UMR5292, Brain
Dynamics and Cognition Team, Lyon F-69000, France 5Institut Universitaire en Santé Mentale de Québec, Université Laval,
Québec, Canada 6Swiss Center for Affective Sciences (CISA), University of Geneva, CH-1205 Geneva, Switzerland and 7Faculté de
Psychologie et des Sciences de l’Education FPSE, University of Geneva, CH-1205 Geneva, Switzerland
Address correspondence to Vincent Rochas, Functional Brain Mapping Laboratory, Department of Fundamental Neuroscience, University of Geneva,
CMU, Rue Michel-Servet 1, 1206 Geneva, Switzerland. Email: [email protected]
V.R. and L.G. contributed equally to this work
ABSTRACT
It has been suggested that the left pre-supplementary motor area (pre-SMA) could be implicated in facial emotion expression
and recognition, especially for laughter/ happiness. To test this hypothesis, in a single-blind randomized crossover study, we
investigated the impact of transcranial magnetic stimulation (TMS) on performances of 18 healthy participants during a facial
emotion recognition task. Using a neuronavigation system based on T1 weighted-MRI of each participant, TMS (5 pulses, 10
Hz) was delivered over the pre-SMA or the vertex (control condition) in an event-related fashion after presentation of happy, fear
and
angry
faces.
Compared to performances during vertex stimulation, we observed that applied over the left pre-SMA, TMS specifically
disrupted facial happiness recognition. No difference was observed between the two conditions neither for fear and anger
recognition, nor for reaction times. Thus, interfering with pre-SMA activity with event-related TMS after stimulus presentation
produced a selective impairment in the recognition of happy faces. These findings provide new insights into the functional
implication of the pre-SMA in facial happiness recognition, which may rely on the mirror properties of pre-SMA neurons.
Keywords: facial emotion recognition, happiness, mirror neurons, pre-SMA, transcranial magnetic stimulation.
107
INTRODUCTION
Facial expressions are a key feature for communication within and across species (Darwin 1872). This type of nonverbal communication is
based on the expression, perception and recognition of facial emotions. Facial emotion recognition (FER) appears to be closely related to
social functioning (Hooker and Park 2002; Addington et al. 2006), and is impaired in a variety of psychiatric and neurological conditions,
such as in schizophrenia (Bediou et al. 2005a, 2007), major depressive disorder (Bediou et al. 2005b), Parkinson’s disease (LachenalChevallet et al. 2006), as well as fronto-temporal dementia and Alzheimer’s disease (Bediou et al. 2009a). FER is also impaired in healthy
individuals with heightened risk for developing schizophrenia (Bediou et al. 2009a). Hence, better understanding the neural mechanisms
implicated in FER is of primary importance.
Numerous imaging studies have investigated the cerebral networks implicated in FER. It has been suggested that complex expressions that
contain blends of emotions may be fully recognized by simulating the perceived expression, either overtly or covertly, and sensing the
emotion produced by that simulation (Adolphs 2001). Thus, FER mechanisms may involve both facial mimicry (i.e. the motor simulation of
another’s expression), as well as empathy (i.e. the sensory simulation of the feelings associated with another’s emotional expression;
Iacoboni 2009). Motor and somatosensory cortical areas may be involved in the motor components of simulation (e.g. facial mimicry),
whereas the amygdala and insula may be involved in the sensory components of simulation (e.g. empathy), and both may contribute to FER
(van der Gaag et al. 2007).
Consistent with this embodied view of emotion recognition (Niedenthal 2007), lesion of the somatosensory cortex (Adolphs et al. 2000), of
the amygdala (Adolphs et al. 1994; Calder et al. 1996) and of the insula (Sprengelmeyer et al. 1997) produce impairments in facial emotion
recognition, possibly reflecting the role of sensory simulation mechanisms or empathy in FER. A peak of activation in pre-supplementary
motor area (pre-SMA) has been observed during a task of facial emotion observation in healthy subjects (Carr et al. 2003), and both the
recognition and the generation of happy and sad expressions activate the pre-SMA (Seitz et al. 2008), consistent with a role in motor
(mimicry) and sensory (empathy) simulation. The fact that the neural responses to the observation and execution of (dynamic) smiles overlap
in premotor cortex and somatosensory cortex (Hennenlotter et al. 2005) is further consistent with a role for these regions in both the
recognition and the motor (mimicry) simulation of this emotion. Furthermore, electrical stimulation of the left pre-SMA with intracranial
subdural electrodes in two epileptic patients has consistently produced laughter (Fried et al. 1998; Krolak-Salmon et al. 2006). In both
studies, laughter was accompanied by sensation of merriment or mirth, and patients gave different explanation for it each time. In addition,
Krolak-Salmon et al. (2006) recorded intracranial evoked potentials in the same epileptic patient during the presentation of emotional faces.
In two different blocks, the patient had to pay attention to gender or emotion. Between 150 and 450 ms after the presentation of an emotional
face (during both tasks), a selective response to happy facial expression was recorded by the electrode implanted in the left pre-SMA. These
studies suggest that the pre-SMA may participate to FER via a mirror communicative activity involved in both the detection and production
of facial emotional expression, especially happiness/laughter. However, these results were obtained in epileptic brains. Although the preSMA was not a part of the patients’ seizures, functional reorganization cannot be excluded. Further studies in healthy subjects are essential to
conclude to a real and systematic implication of the pre-SMA in FER, and especially in facial happiness recognition (FHR). Because of its
cortical location, non-invasive and reversible inhibition of the pre-SMA with transcranial magnetic stimulation (TMS) may be used during an
FER task to assess its causal implication in FHR.
At the interface between neuropsychology and functional neuroimaging, TMS appears to be a suitable means to non-invasively investigate
cerebral function. According to Faraday’s principle, a brief current flows through the stimulation coil producing transient magnetic field that
penetrates the cranium. As a result of this induced magnetic field, an eddy current occurs in brain, transiently and reversibly perturbing
activity in the affected cortical region. Thus, using a perturb-and-measure approach, TMS gives the opportunity to infer about the necessity
(but not the sufficiency) of the integrity of a particular brain region for a given behavior (Paus 2005; Brunelin et al., 2006). For example,
TMS over the medial prefrontal cortex has been shown to reversibly modify the analysis of facially expressed anger (Harmer et al. 2001),
suggesting a crucial role for this region in the recognition of facial anger. Similarly, TMS over the right occipital face ar ea and TMS over the
face region of right somatosensory cortex (relative to the finger region) have been shown to interfere with the processing of emotional facial
expressions but not facial identity (Pitcher et al. 2008). Conversely, TMS of the right STS impaired the processing of gaze direction without
affecting expression processing (Pourtois et al. 2004). Based on past results and its excellent temporal resolution, an event -related TMS
protocol appears appropriate to investigate the role of the pre-SMA in FER.
108
Our study aims to clarify the role of the pre-SMA in FHR. Given (a) the robust activation of the pre-SMA in response to happy faces (Seitz
et al. 2008; Krolak-Salmon et al. 2006), and (b) the laughter and the merriment sensation elicited by electrical stimulation of the pre-SMA in
epileptic patients (Fried et al. 1998; Krolak-Salmon et al. 2006), we hypothesized that pre-SMA is implicated in FER, especially in the
recognition of happiness.
MATERIALS AND METHODS
Subjects
A total of twenty right-handed (average right-handedness score: 97.10; SD = 4.81 %; Edinburgh Handedness Inventory; Oldfield, 1971)
healthy volunteers (twelve males, eight females) aged between 20 and 34 years (mean age = 24.61; SD = 3.74; and years of education = 17;
SD = 2) were enrolled in this single-blind, randomized crossover study in return for payment (100€). University and graduate students were
recruited through ‘word-of-mouth’ according to the following general non-inclusion criteria, which were evaluated during a medical
interview: (1) a story of neurological issues (e.g. epilepsy), (2) personal or familial psychiatric disorder history (axis I DSM IV), (3)
uncorrected vision, (4) pregnancy, (5) TMS contraindication (e.g., metallic prosthesis, pacemaker) and (6) medication intake. All these
exclusion criteria were evaluated during a personal medical interview with a psychiatrist (Personal medical interviews were undertaken by
psychiatrists, Emmanuel Poulet and Mohamed Saoud). Past and current histories of psychiatric disorders were assessed throughout the
clinical interview using the Mini International Neuropsychiatric Interview semi-standardized evaluation (MINI version 4.4). Familial history
was evaluated at the knowledge of the participant and consultation of hospital records. All participants were naive to the FER task, the TMS
tool, and the presented stimuli. They all gave their written consent after a complete description of the study procedure. A local ethical
committee (CPP Sud-Est IV) had approved study design and consent procedure. Subjects were told that they could withdraw from the study
at any time, and one male subject did so before the TMS protocol (n = 19).
Facial Stimuli
Images were taken from a standard set of facial emotion pictures (Ekman and Friesen 1976). Each stimulus was obtained by morphing two
black and white facial pictures (one neutral and one emotional, in different proportions) from a same identity. Morphing construction permits
the creation of an ecological variation in facial emotional intensity and to test subjects’ performances on various levels of difficulties, thus
providing a more sensitive FER measure than classical tests (Bediou et al. 2009a). Moreover, the task should be neither too hard nor too easy
to perform to increase the probability of (a) disrupting the psychological process studied and (b) inferring a functional implication of the
cerebral area stimulated in the evaluated function. In our study, morphed faces were generated between a neutral face and a h appy, angry or
fearful face, for eight identities (four men and four women). The choice of the relevant morphing proportions was based on the results of a
pilot study in an independent sample of 8 healthy volunteers without TMS. Seven levels of morphing between each emotional and neutral
face were used. Our pilot data showed that anger was recognized with greatest difficulty. As a consequence, we used a greater proportion of
the expressive face in each anger morphing (20, 30, 40, 50, 60, 70 and 80% for fear and happiness; 30, 40, 50, 60, 70, 80 and 90% for anger),
in order to keep task difficulty equal between emotions across intensity levels. Pilot data showed that with these morphing levels, recognition
accuracy did not differ significantly between happiness (71.33; SD = 6.53), fear (65.07; SD = 12.72), and anger (66.33; SD = 12.76), F =
1.19, P = 0.34. The total set of stimuli comprised 168 faces (8 identities x 3 emotions x 7 morphs), the order of which was randomized
between series and across subjects. The digitized size, brightness, and contrast of images were standardized.
FER task
FER task was run by the software Presentation v0.55 (Neurobehavioral Systems Inc, Canada), which presented the different images,
recorded subjects’ responses, and controlled the TMS device connected to the computer. Figure 1A depicts the timeline of an experimental
trial. Each trial started with a central white fixation cross on a dark background (duration: 1000, 1250, 1500 or 1750 ms, randomly selected)
attracting subjects’ attention while minimizing response anticipation and motor preparation. Then, a facial stimulus appeared during 50 ms
on a black background. Each stimulus was followed by a black screen (500 ms) during which event-related TMS was applied (1 pulse every
100 ms). This was in turn followed by the response screen (2000 ms). Subjects were instructed to maintain fixation, visualize each picture
109
and judge as quickly and as accurately as possible whether the target face expressed happiness, fear, anger or neutrality (no emotion), by
pressing one of four possible response buttons (up, down, left and right arrow) with their right hand fingers.
Figure 1. (A) Timeline of a TMS experimental trial. Each stimulus appeared during 50 ms and ±1400 ms after crosshair fixation. After stimulus offset, 5 single
pulses of TMS were applied during a black screen (500 ms). The subjects were instructed to answer as quickly and as accurately as possible once the response
screen appeared (2000 ms or until response). Responses were followed by a black screen (500 ms), preceding the next trial. (B) Crossover TMS stimulation
protocol. During the first session, a subject received the TMS pulses over the vertex or the left pre-SMA, and inversely during the second session. The two
sessions were separated by a washout period of 15 days. Short breaks (5-10 min depending on the subject’s comfort) were inserted between each series to
avoid fatigue and to prevent overheating of the stimulator. The order of the TMS conditions was randomized between subjects.
Experimental procedure
In order to standardize experimental conditions, all participants were seated in a padded armchair at a 60 cm distance from the 17-inch
computer monitor, with their head held in place comfortably by a headrest. The subject’s resting motor threshold (RMT) was determined and
study started by a practice block with the same task but with different stimuli (other identities) and without TMS, in order to familiarize with
the task and procedure. Thus, FER under TMS treatment was measured on four occasions: two sessions (pre-SMA, vertex), each comprising
two series of 168 faces. The 168 faces were randomized. Faces were the same for each series and each session. For the two sessions, once the
participant had completed the practice block, the FER task started with a first series of 168 stimuli, lasting about 10 minutes. Then, a second
series of the same 168 faces was shown while TMS was applied on the same site as for the first series. Short breaks (5-10 min depending on
subject’s comfort) were inserted between each series to avoid fatigue and prevent overheating of the stimulator. Thus, for each TMS session,
FER was divided in two series. Because low-frequency repeated TMS (rTMS) research demonstrated delayed and extended effects in time on
several indices of emotional processing (van Honk et al. 2002), the two sessions were separated by 15 days. During the first session, subject
received TMS pulses over the vertex or the left pre-SMA, and inversely during the second session (Figure 1B). The order of TMS stimulation
sites (vertex - pre-SMA or pre-SMA - vertex) was randomized between participants. This had two main advantages. First, it allowed us to
110
compare performance in the two TMS conditions within the same subjects. Second, it allowed us to control for any potential or der/training
effect.
TMS procedure
Neuronavigation
We localized the two TMS sites using a frameless stereotaxic system (Softaxic Optic; http://www.softaxic.com/) to guide TMS coil
positioning over the brain, by means of individual high-resolution T1-weighted MRI transformed in the Talairach space. We targeted sites
based on the Talairach coordinates (x, y, z) for either the left pre-SMA (-6, 15, 58) or the vertex (0, -30, 70). Once MRI co-registration and
cortical target localization were successful, infrared tracking was used to monitor the position of the coil with respect to the participant’s
brain.
Resting motor threshold (RMT)
The RMT was determined as the minimal intensity of electromagnetic stimulation that produced a visible inch adductor contraction in at least
five times out of ten TMS pulses. A figure-8 coil was placed over the participant’s left motor cortex hand area with the coil held tangentially
to the skull and the handle-pointing posterior and down. Single pulses were delivered to the motor cortex, with the intensity of the
stimulation adjusted until a muscle movement in the right hand was visually observed. The location of the stimulation was adjusted to locat e
the inch adductor. Furthermore, it has been demonstrated that rTMS delivered to the primary motor cortex (M1) produces intensitydependent increases in brain activity locally and has associated effects in distant sites with known connection (Speer et al. 2003). In order to
be the more accurate over the pre-SMA and limit the impact of TMS over the network interconnected to pre-SMA, the intensity of TMS was
fixed to 80% RMT which is known to produce a more focal effect (Wagner T. et al. 2009). In addition, a moderate intensity of stimulation
tends to limit the peripheral discomforts and muscles contraction.
TMS pulses
All TMS pulses were delivered by a MagPro X100 magnetic stimulator (MagVenture, Danemark) with a 70mm figure-8 coil at 80%
intensity of each subject’s RMT. Stimulations were controlled through Presentation v0.55 software installed on a computer connected to the
stimulator. Based on previous work (Krolak-Salmon et al. 2006) event-related stimulations were delivered in trains of five pulses (1 pulse
every 100 ms) during the 500 ms after the picture presentation (i.e. the first pulse was synchronized with vertical offset). Each participant
received 3360 pulses during the whole protocol, which lasted approximately 2h (168 stimuli x 5 pulses x 2 series x 2 sessions; 1680 pulses
during each session, separated by two weeks).
Subjective ratings
TMS can perturb subjects’ mood if used daily and repeatedly (Brunelin et al. 2007). Thus, subjects were asked to report their mood on the
Norris’ 16-item visual analog scale (VAS; Norris 1971) before, between and after each series of each session (three measures for each
session).
Data analysis
Post experimental coil positioning
Post experimental visual inspection of the coil localization was conducted to ensure that the stimulation site corresponded t o the desired one.
Data from one female subject had to be excluded from statistical analysis because the coil localization was substantially different at the end
compared to the beginning of the session (n = 18; one subject having withdrawn before the TMS protocol). As a result, statistical analyses
were conducted on data from 18 participants: 10 received TMS over the vertex and then the pre-SMA, and 8 received the inverse sequence
(pre-SMA - vertex).
111
Statistical analyzes
To directly test our a priori hypothesis that TMS over the pre-SMA impairs the recognition of happiness selectively, we subtracted data in
the vertex condition from data in the pre-SMA condition (see also Romei et al., 2011 for a similar approach). This procedure cancels any
individual side-effects due to TMS treatment (e.g. sounds, feelings, stress). The obtained (pre-SMA-minus-vertex) value provides a
quantitative measure of the modification in FER induced by pre-SMA stimulation compared to vertex stimulation for each participant. A
positive value indicates an increase for pre-SMA compared to vertex, whereas a negative value indicates a reduction for pre-SMA compared
to vertex. Statistical analyzes were performed on these pre-SMA-minus-vertex difference scores for both performance (i.e. difference in
percent correct responses, delta-P) and reaction time (i.e. difference in ms between the response screen onset and the onset of the subject’s
response, delta-RT). Only correct trials were considered in the analysis of RT. Considering that FER accuracy was equal across emotions in
our pilot study without TMS stimulation, we then tested whether the delta-P and delta-RT differed significantly from zero, using 2-tailed
one-sample Student t-tests with a significance threshold at P = 0.05 with Bonferroni correction. We predicted a significant change in FHR in
the pre-SMA compared to the vertex condition resulting in a delta-P significantly different from zero for happiness but not for fear and anger.
To investigate the impact of TMS on mood, VAS ratings before and after the session were compared using a repeated measures MANOVA
with the within-subject factors Session (pre vs. post stimulation) and TMS (pre-SMA vs. vertex), and Items (the various aspects rated) as
multiple dependent variable. Considering the risk of low statistical power because of the limited sample size, we therefore averaged the 16
items into an overall mood score and submitted this value to repeated-measure ANOVA examining the impact of TMS and Session only as a
double-check.
To assess the adequacy of our crossover design (and rule-out any possible order/training effect), task performance in the first and second
sessions (all emotions, morphings and TMS conditions collapsed) were compared using 2-tailed paired Student t-test. To quantify a possible
training effect within each session, task performance for the first and the second series of each session (all emotions, morphings and TMS
conditions collapsed) were compared using 2-tailed paired Student t-test.
RESULTS
Effects of TMS on FER
Preliminary considerations
No statistically significant difference was highlighted between performance in the first session compared to the second, t(17) = -0.04, P =
0.97, and between performance in the first compared to the second series of each session, t(17) = -1.21, P = 0.24. In the absence of order
effect (i.e. no difference in FER performance between session 1 and session 2), training effect (i.e. no difference in FER performance
between series 1 and series 2 of each session) and TMS effect on mood (i.e. no difference in VAS ratings before and after TMS), subsequent
analyses examined the impact of TMS on FER data (pre-SMA-minus-vertex) collapsed across series and across TMS conditions, irrespective
of stimulation sequence order (vertex - pre-SMA or pre-SMA - vertex). Although our study was not designed to test gender differences in
FER, or in the effect of TMS on FER, exploratory (i.e. uncorrected) analyses revealed significant differences in FER between men (n = 10,
mean = 54.73 , SD = 5.75) and women (N=8, mean = 63.04, SD = 8.04), t(16) = 2.56, P = 0.021, in line with previous studies (e.g. Montagne
et al. 2005). Significant gender differences in FER were found for happiness and fear in the pre-SMA condition, and for fear in the vertex
condition, but not for anger. Overall accuracy also differed between men and women, in both the vertex and the pre-SMA condition (see
supplementary Table 1). Critically, however, there was no gender difference in the effect TMS on FER when pre-SMA data were subtracted
from vertex data (see supplementary Table 2).
Effects of TMS on FER performance
On average (i.e. all emotions and morphings collapsed), participants’ performance was 57.62% (SD = 10.54%) in the pre-SMA condition and
59.24% (SD = 11.30%) in the vertex condition, a difference that was statistically significant, t(17) = -2.20 , P = 0.04. As reported in Table 1
and illustrated in Figure 2, the delta-P for happiness differed significantly from zero, t(17) = 2.95, P = 0.009, whereas the delta-P for fear
and anger did not, t(17) = 0.56, P = 0.59; t(17) = 0.53, P = 0.60, respectively.
112
Table 1. The mean score (expressed in percentage of correct responses) for the different emotions (happiness, fear, and anger, all morphings collapsed) in the
vertex and pre-SMA conditions, delta-P (%; pre-SMA − vertex) and 2-tailed 1-sample Student t-test (P values; n = 18) all morphings collapsed. Only the deltaP for happiness differed significantly from zero (P = 0.009)
Figure 2. Delta-P values (percentage of correct responses in the pre-SMA condition minus percentage of correct responses in the vertex condition) for each
subject and each emotion (averaged across all morphings). The delta-P for happiness differed significantly from zero, whereas the delta-P for fear and anger
did not (see text for statistics).
As expected, subjects had more difficulty identifying happiness in the pre-SMA TMS condition than in the vertex TMS condition, whereas
fear and anger recognition were not significantly affected by TMS over the left pre-SMA compared to the TMS over the vertex.
Effects of TMS on RTs
RT data for each emotion and TMS condition (i.e., collapsed across all morphs), as well as the pre-SMA-minus-vertex difference in RT
(delta-RT) are summarized in Table 2. None of the delta-RT value differed significantly from zero, suggesting that TMS did not affect RTs.
113
Table 2. Mean and SD for RT (ms) for correct trials for the different emotions (happiness, fear, and anger, all morphings collapsed) in the vertex and pre-SMA
conditions, delta-RT (ms; pre-SMA − vertex), and 2-tailed 1-sample Student t-test (P values; n = 18). No statistical difference was highlighted.
Effects of TMS on mood
We found no effect of TMS on mood. A repeated measures MANOVA with the within-subject factors Session (pre vs. post stimulation) and
TMS (pre-SMA vs. vertex) and Items as multiple dependent variable yielded a significant main effect of Item, F = 280, P < 0.001.
Importantly, however, there was no significant effect or interaction with the factors Session and TMS (all F’s < 1), suggesting that our TMS
protocol did not significantly affect participants’ mood. We note, however, that running this analysis with our limited sample size bears the
risk of low statistical power. We therefore averaged the 16 items into an overall mood score and submitted this value to repeated-measure
ANOVA examining the impact of TMS and Session. This analysis showed very similar results; there were no significant main effect of
TMS, and no TMS x Session interaction (all F’s < 1). A marginal effect of session, F = 3.47, P = 0.08, suggested that mood varied between
the beginning and the end of each session, probably due to fatigue. Critically though, this effect was not affected by the TMS condition. VAS
ratings before and after each TMS session did not differ significantly (before pre-SMA: mean = 51.57, SD = 34.17; after pre-SMA: mean =
52.17, SD = 33.92; before versus after pre-SMA: t(15) = -0.44, P = 0.67; before vertex: mean = 50.54, SD = 32.21; after vertex: mean =
51.68, SD = 33.99, before versus after vertex: t(15) = -1.14, P = 0.27), suggesting that subjects’ mood was not affected by TMS. There was
no significant correlation between FER and mood (all R’s < 0.34, P’s > 0.17).
In summary, compared to TMS over the vertex, TMS over the pre-SMA impaired selectively the recognition of happy facial expressions,
without affecting the recognition of anger and fear, and without affecting RTs and mood.
DISCUSSION
The primary goal of the present study was to assess, using an interference technique (TMS), whether the pre-SMA is involved in FHR. We
hypothesized that compared to TMS over the vertex, TMS over the left pre-SMA would specifically interfere with the recognition of
happiness. As predicted, we showed that TMS over left pre-SMA impaired the recognition of happy faces, without affecting the recognition
of fearful and angry faces, and without affecting RT. There was no evidence that TMS pulses delivered during this study led to undesirable
short- and long-term effects, and none of the subjects included in our study reported adverse event. Moreover, we found no effect of TMS on
mood, and no relationship between mood and FER. Hence, reduced happiness recognition following TMS stimulation of the left pre-SMA
compared to the vertex, may be attributed to the perturbation of neural activity in the pre-SMA or in a broader neural network including this
structure. Although the precise mechanism(s) by which pre-SMA may be involved in the recognition of facial happiness remains unclear, our
study provides the first evidence for a direct relationship between pre-SMA activity and recognition accuracy for happy faces in healthy
subjects.
Previous studies examining the impact of repetitive TMS stimulation (rTMS) of lateral prefrontal cortical areas (PFC) on mood suggest that
the effects are opposed depending on the hemisphere stimulated (left vs. right) and on the frequency of stimulation (low frequency, LF vs.
high frequency, HF). In healthy volunteers, left PFC HF stimulation increases self-rated sadness (George et al. 1996; Pascual-Leone et al.
1996b; Dearing 1997), whereas HF stimulating of the right PFC increases self-rated happiness (e.g. George et al. 1996; Pascual-Leone et al
1996b), though negative results have also been reported (Mosimann et al. 2000). However, rTMS has been used successfully to treat
114
depressive symptoms in patients with major depressive disorder with two main approaches: HF rTMS of the left dorsolateral prefrontal
cortex or LF rTMS of the right dorsolateral prefrontal cortex (George et al. 1999; Klein et al. 1999; Post et al. 1999; Eche et al. 2012). These
therapeutic effects are confirmed by several large-scale clinical trials and a number of meta-analyses (see Padberg & George 2009; Fitzgerald
2011 for recent reviews). Interestingly, rTMS also has lateralized effects on facial expressions in depressed patients. In particular, the
frequency of laughter was increased after stimulation of the left PFC and decreased following stimulation of the right PFC (Padberg et al.
2001). In sum, similar effects are found by either stimulating the left prefrontal cortex with high frequency, or inhibiting the right prefrontal
cortex with low frequencies, but opposite effects are found with the same stimulation protocol in depressed patients and healthy controls.
Here, we stimulated a different but connected region (the left pre-SMA) using 5 pulses of event-related TMS at 10 Hz (transient lesion) and
found no effect on mood, making it unlikely that the TMS-induced perturbation of happiness recognition is an indirect consequence of the
impact of TMS on mood. Although the disruption of activity in the left pre-SMA impaired the recognition of happiness without any short
term effect on mood, it is plausible that the modification in FER – in this case in happiness recognition – would affect mood in the long term
(e.g. with a prolonged rTMS treatment), similar to what is observed following antidepressants administration in both healthy volunteers
(Harmer et al. 2004) and depressed patients (Harmer et al. 2009a), in which changes in facial expression processing (especially fear and
happiness recognition) are observed several days or weeks before changes in mood or depressive symptoms, and actually predicting these
changes (Harmer et al. 2003; 2009b).
In addition to its effects on mood, rTMS of the dorsolateral PFC has been shown to affect attention and physiological responses in healthy
volunteers (van Honk et al. 2003). LF rTMS of right prefrontal areas reduces attention to (unmasked) fearful faces (van Honk et al. 2002) and
increases attention towards angry faces (d’Alfonso et al. 2001), whereas left rTMS diverts attention away from angry faces. However, the
hemispheric lateralization of HF rTMS effects may depend on additional factors, such as the sex and the valence and/or motivational
direction of the emotional expression (Brüne et al. 2006), though in this study the authors stimulated the left vs. right temporal (not frontal)
cortex and only included healthy female subjects. However, the transient modification of FHR by left pre-SMA TMS cannot be accounted
for a general effect of TMS on attention for at least 2 reasons. First, the disruption was specific to happiness, and second, there were no
differences in RTs between emotions and no effect of TMS on RTs. Thus, we surmise that the decrease in FHR is caused by the impact of
TMS on a selective mimicry-like mechanism involving the mirror properties of the preSMA, as discussed in more details here below. Our
findings extend the current literature on the neurobiology of FER by showing that event-related TMS (as opposed to rTMS) of the left preSMA can impair the recognition of happiness selectively without any short-term modifications of mood and attention. Previous studies
already suggested an implication of the somatosensory cortex in FER (Pourtois et al. 2004; Pitcher et al. 2007, 2008, 2009), and of the medial
PFC in anger (Harmer et al. 2001).
Recent work (Mukamel et al. 2010) suggests that some neurons in the human pre-SMA show mirror properties - i.e. discharging when
executing a given motor act and when observing the same action being performed by someone else. An important element for understanding
the selective impact of pre-SMA stimulation on happiness recognition is the motor aspect of facial emotional expressions. Facial expressions
are differentiated on the basis of the activity of specific facial muscles (Ekman et al. 1978). In particular, anger is characterized by an
increased activity of the Corrugator, producing frowning (Duchenne 1859). Similarly, fear is associated by an increased activity of the
Orbicularis oris (Duchenne 1859) responsible for eyes-opening. Unlike these two expressions involving mainly the eyes region, happiness is
easily recognizable via the contraction of the Zygomaticus characterizing smiles (Duchenne 1859). Happiness is also known to be
particularly contagious (Dimberg et al. 2000). Passive viewing of happy faces induces contractions of the Zygomaticus (Hatfield et al.,
1993), suggesting that this emotion is particularly keen to activate mirror neuron mechanisms. Importantly, the repertoire of the mirror
neuron system indeed extends from hand actions to a wide range of body actions including facial actions (Buccino et al., 2001). Furthermore,
the left SMA (SMA-proper and pre-SMA), but not the right, has a bilateral face representation essential in producing facial expressions
(Fried et al. 1991). Facial happiness expression is intrinsically related to mouth movements, suggesting that pre-SMA mirror neurons may
potentially discharge in relation to mouth movement. Consistent with this idea, increasing the intensity of an emotional expression (i.e.
morphing level) during passive viewing is associated with increases in both the evoked neural activity and the facial muscular activity
involved in the expression of the perceived emotion (Achaibou et al. 2008). Thus, the observed effect of left pre-SMA stimulation on
happiness recognition may be due to an impact of TMS on the activity of pre-SMA mirror neurons involved in the perception and production
of mouth movements, or in their simulation. Just like mirror neurons located in the somatosensory cortex, mirror neurons in the pre-SMA
115
may be involved in embodied cognition, and more specifically in the (motor) simulation mechanisms (e.g., facial mimicry) that are known to
facilitate FER (Niedenthal 2007), and more particularly so for happiness (Oberman et al. 2007). The fact that a significant proportion of
mirror neurons in the pre-SMA respond to communicative mouth movements (Mukamel et al. 2010) brings further support for this
interpretation.
In our study, the five TMS pulses were applied over the left pre-SMA 50, 150, 250, 350 and 450 ms after the offset of the facial stimulus
(thus, between 100 and 500 ms after stimulus onset). Our results are thus consistent with past electrophysiological studies showing a preSMA implication in FHR between 150 and 450 ms after the stimulus onset (Krolak-Salmon et al. 2006) or between 100 and 720 ms after
stimulus onset (Seitz et al. 2008). Current models (e.g. Adolphs 2002) suggest that the information sufficient to distinguish faces from other
objects is encoded within 120 ms, whereas the construction of a detailed perceptual representation of a face requires about 170 ms, and the
conceptual knowledge of the emotion signaled by the face, more than 300 ms. Furthermore, information sufficient to distinguish among
different emotional expressions appears around 170 ms after the onset of the stimulus, suggesting that responses to emotional stimuli in
visual cortices are modulated by a feedback from interconnected structures, such as the amygdala and orbitofrontal cortex (Adolphs 2002),
where rapid responses to facial expressions have been recorded (Kawasaki et al. 2001; Krolak-Salmon et al. 2004). In line with this model,
activity differentiating between specific emotional expressions can be recorded between 250 and 550 ms after stimulus onset b oth
intracranially (Krolak-Salmon et al. 2003; 2004) and on the scalp (Bediou et al. 2006; Krolak-Salmon et al. 2001) and even before over
frontocentral electrodes (Bediou et al. 2009b).
Our results suggest a direct relationship between the activity of the left pre-SMA and FHR. The pre-SMA may react to happy faces very
rapidly (within 100 to 450 ms), most likely via interactions with the orbitofrontal cortex, the amygdala and occipitotemporal areas. The
amygdala and orbitofrontal cortex may generate an emotional response in the subject, via thalamic connections to motor struct ures (e.g. the
pre-SMA; Inase et al. 1996), hypothalamus, and brainstem nuclei, where components of an emotional response to the facial expression can
be activated (Adolphs 2002). This mechanism may contribute to the generation of knowledge about another person’s emotional state, via the
process of simulation by motor mirror neurons, and would draw on somatosensory related cortices in the right hemisphere for representing
the emotional changes in the perceiver (Adolphs 2001; Pitcher et al. 2008). Further studies are needed to uncover the dynamic functional
connectivity of the pre-SMA with other brain areas involved in FHR. Double-pulses of TMS with 50 ms between pulses could be delivered
at different times from stimulus onset, in order to pinpoint the timing, and causality, of pre-SMA implication in FHR.
Because of our cross-over TMS design, we were constrained in the number of trials per subject and thereby in the number of experimental
conditions (i.e. emotions and morphing levels). Various arguments guided our choice of emotional expressions. First, fear and anger differ
from happiness on valence and motivational direction, two of the main underlying dimensions of emotion. Previous studies have found that
the effects of TMS on FER depend on the valence or motivational categories of the emotions considered, and on the lateralization of the
stimulation (Baeken et al. 2011; d'Alfonso et al. 2000). Considering that we were targeting the left pre-SMA to investigate its implication in
happiness recognition (positive valence, approach motivation), our choice of fear (negative valence, avoidance motivation) and anger
(negative valence, approach motivation) was motivated by the existence of the competing theories about the lateralization of emotions
(Davidson 2004 and Harmon-Jones 2004). Moreover, fear and anger are known to attract strong attention, and together with happiness, are
generally recognized easier than other negative emotions, such as disgust or sadness. Although sadness would have been the most intuitive
emotion to oppose to happiness, this emotion tends to be poorly recognized in FER studies, especially when using morphed faces (Montagne
et al. 2007). In addition, the neural circuitry underlying the perception and recognition of fear and anger is at least partly established, whereas
the neurobiology of sadness recognition is much less clear. The neural basis of disgust recognition is also partly known, but when used with
anger, the two emotions are less recognized. Thus, the fact that we observed a significant impairment in FHR following the di sruption of the
left pre-SMA activity is further consistent with an involvement of the left PFC in the recognition of a positive valence, and approach-related,
emotional expression.
Our control condition (TMS over the vertex) may be subject of controversy. An appropriate sham should stimulate the ancillary aspects of
TMS, such as scalp stimulation and acoustic artifacts, as closely as possible to experimental TMS, but should not result in cortical
stimulation. Available sham coils fail to truly mimic the peripheral sensations associated with TMS easily, such that it becomes obvious to all
subjects in a crossover protocol whether they are receiving the real or placebo stimulation. Furthermore, previous research has shown that
performance and RTs in a FER task were not affected by TMS over the vertex compared to a no-TMS condition (Pitcher et al. 2008). For
116
these reasons, we used the same figure-8 coil over the vertex for our TMS-control condition. To our knowledge, the vertex is an appropriate
control site for TMS stimulation in a FER task (Pitcher et al. 2008) in that it does not interfere with attentional processes, vision and emotion
recognition. Moreover, our VAS analysis showed that subjects’ mood was not affected by TMS over the vertex. In the current study, none of
the subjects was able to say whether he or she was stimulated on the vertex or pre-SMA. Thus, modifications of FHR performance can
reliably be attributed to the functional TMS-induced perturbation of the targeted cortical area which is the only parameter changing between
the two sessions. As expected, TMS over the left pre-SMA resulted in lower performance for happiness recognition. Such an emotionspecific impairment is compatible with a selective involvement of the left pre-SMA in the processing of facial expressions of happiness
(Krolak-Salmon et al. 2006).
In conclusion, we have demonstrated that the functional integrity of the left pre-SMA is indispensable for the recognition of happy but not
angry and fearful faces. The present research provides new insights into the functions of this region and provides the first direct link between
the activity of this region and the performance in a social cognitive task. Combined to works disclosing selective pre-SMA reaction to happy
faces, the present study supports the existence of mirror properties of pre-SMA neurons, which may represent a neural basis of embodied
FER mechanisms that create a direct link between the sender and the receiver of a social message.
FUNDING
This research was supported by two grants from CSR (Scientific Council of Research; CSR 2006 and 2010) of Le Vinatier Hospital.
REFERENCES
Achaibou A, Pourtois G, Schwartz S, Vuilleumier P. 2008. Simultaneous recording of EEG and facial muscle reactions during spontaneous
emotional mimicry. Neuropsychologia. 46(4):1104-13.
Addington J, Saeedi H, Addington D. 2006. Facial affect recognition: A mediator between cognitive and social functioning in psychosis?
Schizophrenia Research. 85:142-50.
Adolphs R. 2001. The neurobiology of social cognition. Current Opinion in neurobiology. 11:231-9.
Adolphs R. 2002. Neural systems for recognizing emotion. Current Opinion in Neurobiology. 12:169-77.
Atkinson AP, Adolphs R. 2011. The neuropsychology of face perception: beyond simple dissociations and functional selectivity.
Philosophical transactions of the Royal Society of London. Series B, Biological sciences, 366(1571), 1726-38.
Baeken C, Van Schuerbeek P, De Raedt R, De Mey J, Vanderhasselt MA, Bossuyt A, Luypaert R. 2011. The effect of one left-sided
dorsolateral prefrontal sham-controlled HF-rTMS session on approach and withdrawal related emotional neuronal processes. Clin
Neurophysiol. 122:2217-2226.
Bediou B, Franck N, Saoud M, Baudouin JY, Tiberghien G, Dalery J and d’Amato T. 2005a. Effects of emotion and identity on facial affect
processing in schizophrenia. Psychiatry research. 133(2-3):149-57.
Bediou B, Krolak-Salmon P, saoud M, Henaff MA, Burt M, Dalery J, d’Amato T. 2005b. Facial expression and sex recognition in
schizophrenia and depression.Can J Psychiatry. 50(9):525-33.
Bediou B, Henaff MA, Bertrand O, Brunelin J, d’Amato T, Saoud M, Krolak-Salmon P. 2007 Impaired fronto-temporal processing of
emotion in schizophrenia. Clinical Neurophysiology. 37(2):77-87.
Bediou B, Ryff I, Milliery M, Hénaff MA, d’Amato T, Bonnefoy M, Vighetto A, Krolak-Salmon P. 2009a Impaired social cognition in mild
Alzheimer disease. J Geriatr Psychiatry Neurol. 22(2):130-40.
Bediou B, Eimer M, d'Amato T, Hauk O, Calder AJ. 2009b. In the eye of the beholder: individual differences in reward-drive modulate early
frontocentral ERPs to angry faces. Neuropsychologia. 47:825-834.
Brüne M, Bahramali H, Hennessy M, Snyder A. 2006. Are angry male and female faces represented in opposite hemispheres of the female
brain? A study using repetitive transcranial magnetic stimulation (rTMS). J Integr Neurosci. 5:187-197.
Brunelin J, Poulet E, Bediou B, Kallel L, Dalery J, D'amato T, Saoud M. 2006. Low frequency repetitive transcranial magnetic stimulation
improves source monitoring deficit in hallucinating patients with schizophrenia. Schizophr Res. 81(1):41-5
Brunelin J, Poulet E, Boeuve C, Zeoug-vial H, d’Amato T and Saoud M. 2007 Efficacy of repetitive transcranial magnetic stimulation
(rTMS) in major depression: a review. Encephale. 33(2):126-34.
117
Buccino G, Binkofski F, Fink GR, Fadiga L, Fogassi L, Gallese V, Seitz RJ, Zilles K, Rizzolatti G, Freund HJ. 2001. Action observation
activates premotor and parietal areas in a somatotopic manner: an fMRI study. European Journal of Neuroscience. 13:400-4.
Carr L, Iacaboni M, Dubeau MC, Mazziotta JC, Lenzi GL. 2003. Neural mechanisms of empathy in humans: A relay from neural systems for
imitation to limbic areas. Proc. Natl. Acad. Sci. U. S. A. 100:5497-502.
D'Alfonso AAL , van Honk J, Hermans E, Postma A, de Haan EHF. 2000. Laterality effects in selective attention to threat after repetitive
transcranial magnetic stimulation at the prefrontal cortex in female subjects. 280(3):195-198.
Darwin C. 1872. The expression of emotions in man and animals. New York: Phylosophical Library.
Davidson RJ. 2004. What does the prefrontal cortex “do” in affect: perspectives on frontal EEG asymmetry research. Biological Psychology.
67(1-2):219-234.
Dearing Martin J, George MS, Greenberg BD, Wassermann EM, Schlaepfer TE, Murphy DL, Hallett M, Post RM. 1997. Mood effects of
prefrontal repetitive highfrequency TMS in healthy volunteers. CNS Spectrums. 2:53-68.
Dimberg U, Thunberg M, Elmehed K. 2000. Unconscious facial reactions to emotional facial expressions. Psychological Science. 11(1), 86–
89.
Duchenne GBA. 1859. The mechanism of human facial expression. New York: Cambridge University Press.
Eche J, Mondino M, Haesebaert F, Saoud M, Poulet E, Brunelin J. 2012. Low- vs high-frequency repetitive transcranial magnetic stimulation
as an add-on treatment for refractory depression. Front Psychiatry. 3: 13.
Ekman P, Friesen WV. 1976. Pictures of facial affect. Palo Alto (CA): Consulting Psychologists Press.
Ekman P, Friesen WV. 1978. Facial Action Coding System: A Technique for the Measurement of Facial Movement. Palo Alto (CA):
Consulting Psychologists Press.
Fitzgerald, PB, Daskalakis ZJ. 2001 The effects of repetitive transcranial magnetic stimulation in the treatment of depression. Expert Review
of Medical Devices. 8(1): 85-95(11).
Fried I, Katz A, McCarthy G, Sass KJ, Williamson P, Spencer SS, Spencre DD. 1991. Functional organization of human supplementary
motor cortex studied by electrical stimulation. The Journal of Neuroscience. 11(11):3656-66.
Fried I, Wilson CL, MacDonald KA, Behnke EJ. 1998. Electric current stimulates laughter. Nature. 391:650.
George MS, Wassermann EM, Williams WA, Steppel J, Pascual-Leone A, Basser P, Hallett M, Post RM. 1996. Changes in mood and
hormone levels after rapid-rate transcranial magnetic stimulation (rTMS) of the prefrontal cortex. J Neuropsychiatry Clin Neurosci.
8:172-180.
George MS, Lisanby SH, Sackeim HA. 1999. Transcranial magnetic stimulation. Arch Gen Psychiatry. 56:300-311.
Harmer CJ, Thilo KV, Rothwell JC, Goodwin GM. 2001. Transcranial magnetic stimulation of medial-frontal cortex impairs the processing
of angry facial expressions. Nat Neurosci. 4:17-8.
Harmer CJ, Hill SA, Taylor MJ, Cowen PJ, Goodwin GM. 2003. Toward a neuropsychological theory of antidepressant drug action: increase
in positive emotional bias after potentiation of norepinephrine activity. Am J Psychiatry. 160:990-992.
Harmer CJ, Shelley NC, Cowen PJ, Goodwin GM. 2004. Increased positive versus negative affective perception and memory in healthy
volunteers following selective serotonin and norepinephrine reuptake inhibition. Am J Psychiatry. 161:1256-1263.
Harmer CJ, O'Sullivan U, Favaron E, Massey-Chase R, Ayres R, Reinecke A, Goodwin GM, Cowen PJ. 2009a. Effect of acute
antidepressant administration on negative affective bias in depressed patients. Am J Psychiatry. 166:1178-1184.
Harmer CJ, Goodwin GM, Cowen PJ. 2009b. Why do antidepressants take so long to work? A cognitive neuropsychological model of
antidepressant drug action. Br J Psychiatry. 195:102-108.
Harmon-Jones E. 2004. Contributions from research on anger and cognitive dissonance to understanding the motivational functions of
asymmetrical frontal brain activity. Biological Psychology. 67(1-2):51-76.
Hatfield E, Cacciopo J, Rapson RL. 1993 Emotional contagion. Current Directions in Psychological Sciences. 2:96-9.
Hennenlotter A, Schroeder U, Erhard P, Castrop F, Haslinger B, Stoecker D, Lange KW, Caballos-Baumann AO. 2005. A common neural
basis for receptive and expressive communication of pleasant facial affect. NeuroImage. 26:581-591.
118
Hooker C, Park S. 2002. Emotion processing and its relationship to social functioning in schizophrenia patients. Psychiatry Research.
112:41-50.
Iacoboni M. 2009. Imitation, Empathy, and Mirror Neurons. Ann Rev Psychol. 60:653-70.
Inase M, Tokuno H, Akazawa T, Takada M. 1996. Origin of thalamocortical projections to the presupplementary motor area (pre-SMA) in
the macaque monkey. Neuroscience Research. 5:217-27.
Kawasaki H, Adolphs R, Kaufman O, Damasio H, Damasio AR, Granner M, Bakken H, Hori T, Howard MA. 2001. Single-unit responses to
emotional visual stimuli recorded in human ventral prefrontal cortex. Nature Neuroscience. 4:15-16.
Klein E, Kreinin I, Chistyakov A, Koren D, Mecz L, Marmur S, Ben-Shachar D, Feinsod M. 1999. Therapeutic efficacy of right prefrontal
slow repetitive transcranial magnetic stimulation in major depression. Arch Gen Psychiatry. 56:315-320.
Krolak-Salmon, P., Fischer, C., Vighetto, A., Mauguiere, F. 2001. Processing of facial emotional expression: spatio-temporal data as
assessed by scalp event-related potentials. Eur J Neurosci. 13(5):987-994.
Krolak-Salmon P, Henaff MA, Isnard J, Tallon-Baudry C, Guenot M, Vighetto A, Bertrand O, Mauguiere F. 2003. An attention modulated
response to disgust in human ventral anterior insula. Ann Neurol. 53:446-453.
Krolak-Salmon P, Henaff MA, Vighetto A, Bertrand O, Mauguiere F. 2004. Early amygdala reaction to fear spreading in occipital, temporal,
and frontal cortex: a depth electrode ERP study in human. Neuron. 42:665-676.
Krolak-Salmon P, Hénaff MA, Vighetto A, Bauchet F, Bertrand O, Mauguière F, Isnard J. 2006. Experiencing and Detecting Hapiness in
Humans: The Role of the Supplementary Motor Area. Ann. Neurol. 59:196-99.
Lachenal-Chevallet K, Bediou B, Bouvard M, Thobois S, Broussole E, Vighetto A, Krolak-Salmon P. 2006 Emotional facial expression
recognition impairment in Parkinson disease. Psychol Neuropsychiatr Vieil, 4(1):61-7.
Lee TW, Josephs O, Dolan RJ, Critchley HD. 2006. Imitating expressions: emotion-specific neural substrates in facial mimicry. Soc Cogn
Affect Neurosci. 1:122-135.
Montagne B, Kessels RP, Frigerio E, de Haan EH, Perrett DI. 2005. Sex differences in the perception of affective facial expressions: do men
really lack emotional sensitivity? Cogn Process. 6:136-141.
Mosimann UP, Rihs TA, Engeler J, Fisch H, Schlaepfer TE. 2000. Mood effects of repetitive transcranial magnetic stimulation of left
prefrontal cortex in healthy volunteers. Psychiatry Res. 94:251-256.
Mukamel R, Ekstrom AD, Kaplan J, Iacaboni M, Fried I. 2010. Single-neuron responses in humans during execution and observation of
actions. Current Biology. 20:750-6.
Niedenthal PM. 2007 Embodying emotion. Science. 316:1002-5.
Norris H. 1971. The action of sedatives on brain Stem oculomotor systems in man. Neuropharmacology. 10:181-91.
Oberman LM, Winkelman P and Ramachandran VS. 2007. Face to face: blocking facial mimicry can selectively impair recognition of
emotional expressions. Social Neuroscience. 2(3-4):167-78.
Oldfield RC. 1971. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia. 9:97-113.
Padberg F, Juckel G, Prassl A, Zwanzger P, Mavrogiorgou P, Hegerl U, Hampel H, Moller HJ. 2001. Prefrontal cortex modulation of mood
and emotionally induced facial expressions: a transcranial magnetic stimulation study. J Neuropsychiatry Clin Neurosci. 13:206-212.
Padberg F, George MS. 2009. Repetitive transcranial magnetic stimulation of the prefrontal cortex in depression. Experimental Neurology.
219:2–13.
Pascual-Leone A, Catala MD, Pascual-Leone Pascual A. 1996. Lateralized effect of rapid-rate transcranial magnetic stimulation of the
prefrontal cortex on mood. Neurology. 46:499-502.
Paus T. 2005. Inferring causality in brain images: a perturbation approach. Philos. Trans. R. Soc. Lond., B, Biol. Sci. 360:1109-14.
Pitcher D, Walsh V, Yovel G, Duchaine B. 2007. TMS evidence for the involvement of the right occipital face area in early face processing.
Curr Biol. 17:1568-1573.
Pitcher D, Garrido L, Walsh V, Duchaine BC. 2008. Transcranial magnetic stimulation disrupts the perception and embodiment of facial
expressions. The Journal of Neuroscience. 28:8929-33.
Pitcher D, Charles L, Devlin JT, Walsh V, Duchaine B. 2009. Triple dissociation of faces, bodies, and objects in extrastriate cortex. Curr
Biol. 19:319-324.
119
Pourtois G, Sander D, Andres M, Grandjean D, Reveret L, Olivier E, Vuilleumier P. 2004. Dissociable roles of the human somatosensory
and superior temporal cortices for processing social face signals. Eur J Neurosci. 20:3507-3515.
Post RM, Kimbrell TA, McCann UD, Dunn RT, Osuch EA, Speer AM, Weiss SR. 1999. Repetitive transcranial magnetic stimulation as a
neuropsychiatric tool: present status and future potential. J ECT. 15:39-59.
Seitz RJ, Schäfer R, Scherfeld D, Friederichs S, Popp K, Wittsack HJ, Azari NP, Franz M. 2008. Valuating other people’s emotional face
expression: a combined functional magnetic resonance imaging and electroencephalography study. Neuroscience. 152:713-22.
Speer AM, Willis MW, Herscovitch P, Daube-Witherspoon M, Shelton JR, Benson BE, Post RM, Wassermann EM. 2003. Intensitydependent regional cerebral blood flow during 1-Hz repetitive transcranial magnetic stimulation (rTMS) in healthy volunteers studied
with H215O positron emission tomography: I. Effects of primary motor cortex rTMS. Biological Psychiatry. 54(8):818-25.
Van der Gaag C, Minderaa RB, Keysers C.2007. Facial expressions: What the mirror neuron system can and cannot tell us. Social
Neuroscience. 2:179-222.
Van Honk J, Schutter JLG, d’Alfonso AAL, Kessels RPC, de Haan EHF. 2002. 1 Hz rTMS over the Right Prefrontal Cortex Reduces
Vigilant Attention to Unmasked but Not to Masked Fearful Faces. Society of Biological Psychiatry. 52:312-17.
Van Honk J, Schutter DJ, Putman P, de Haan EH, d'Alfonso AA. 2003. Reductions in phenomenological, physiological and attentional
indices of depressive mood after 2 Hz rTMS over the right parietal cortex in healthy human subjects. Psychiatry Res. 120:95-101.
Wagner T, Rushmore J, Eden U & Valero-Cabre A. 2009. Biophysical foundations underlying TMS: setting the stage for an effective use of
neurostimulation in the cognitive neurosciences. Cortex. 45:1025–1034.
120
2. The second study using EEG and TMS: on lexical detection of emotional words
Exp Brain Res (2014) 232:1267-1281
DOI 10.1007/s00221-014-3843-y
Research Article
Very early processing of emotional words revealed in temporoparietal
junctions of both hemispheres by EEG and TMS
Vincent Rochas · Tonia A. Rihs · Nadia Rosenberg ·
Theodor Landis · Christoph M. Michel
Received: 27 August 2013 / Accepted: 10 January 2014 / Published online: 5 February 2014
© Springer-Verlag Berlin Heidelberg 2014
ABSTRACT
We investigate the contribution of both hemispheres in a lateralized lexical decision paradigm with emotional and neutral
words in healthy volunteers. In a first experiment, high density EEG analysis using source imaging methods revealed early
specific participation of the temporoparietal junctions in both hemispheres for the detection of words. Then in an event
related TMS experiment with the same task, the disruption of left or right temporoparietal junctions compared to a control
stimulation over the vertex showed a slowing that is more pronounced when words are emotional and presented in the LVF.
This indicates that interference with both left and right temporoparietal junctions results in impaired processing of words that
were presented to the left visual field. In addition, these results point to a specific cooperative contribution of the right
hemisphere in the processing of words with emotional content compared to neutral words at very early stages. Results from
the two experiments can be integrated in a brain-based spatiotemporal model of the early detection of written words.
Keywords: Lexical decision; word detection; emotion; lateralization; right hemisphere; EEG; source imaging; TMS.
INTRODUCTION
Since the seminal works of Broca (1861) and Wernicke (1874), the left cerebral hemisphere in right-handers is known to be dominant for
language processing. However the study of split-brain patients with surgically disconnected hemispheres (Sperry 1961; Gazzaniga and
Sperry 1967; Gazzaniga and Hillyard 1971; and Gazzaniga 2005 for a review) and the study of healthy subjects with lateralized stimulus
presentations (for a review see Springer and Deutsch 2004) leave no doubt that the right hemisphere also plays a considerable role in
language processing. More recently, neuropsychological studies, lesion-based studies, and imagery studies (Stroobant et al. 2009;
121
Lidzbaetal. 2011; Mohr B. et al. 2007) revealed the participation of the right hemisphere in various features of language processing – e.g.
intonation (Ross et al. 1988; Pell 2007), concreteness (Papagno et al. 2009) or context-related language (Diaz and Hogstrom 2011). Most
classical is the relation of the right hemisphere with the emotional dimension of language processing. The special role that emotional
words play with respect to cerebral laterality has been recently summarized (Landis 2006).
The present study aims to assess language lateralisation and the interaction between the two hemispheres in relation to the treatment of
the emotional content of the words. It is based on earlier work from our group (Ortigue et al. 2004) that used a lexical decision task with
a simultaneous bilateral presentation of letter strings. In this previous work, event-related potential (ERP) analysis showed a distinct
scalp topography for emotional words presented in the right visual field, indicative of a specific intracranial generator configuration for
these words. The well-known emotional word advantage was accompanied by increased activation of right lateral occipital brain areas at
around 100-140 ms. However, somewhat unexpectedly, this effect was only found for emotional words presented to the right visual field
(RVF) and not to left visual field (LVF), despite the enhanced performance for emotional words in both conditions.
Derived from the study of Ortigue et al. (2004), the present study does not only better control the stimulus material and improve the
analysis of the ERPs, it also assesses the early contribution of the right hemisphere to emotional word processing by applying
Transcranial Magnetic Stimulation (TMS). We performed a first experiment where we tracked the early brain processes of real word
detection by high density electroencephalography (256-channel EEG) using statistical analysis of the ERPs in the sensor space (voltage
topography) and source space (electrical source imaging) (Michel and Murray 2011). In the second experiment on an independent group
of participants, event-related TMS was alternatively delivered in two regions that were identified in the first experiment to be recruited in
these early stages of word processing – the left and right temporoparietal junctions (TPJ). The time window of stimulation was also
selected on the basis of the first experiment. Behavioural measurements after interruption of these areas allowed to evaluate their
implication in word processing.
The combination of these two time-resolved brain imagery methods allows a comprehensive understanding of particular actors and
processes of the network in a given cognitive task. Our findings are then discussed and included in a functional spatiotemporal model of
early processing of written words and emotionality.
MATERIALS & METHODS
Study 1 (EEG)
Participants
Eighteen healthy volunteers participated in this study approved by the Ethical Committee of the University Hospital of Geneva,
Switzerland. Included participants were young adult males, right-handed, aged from 20 to 32 years (mean age= 24.8; SD = 3.5). All had
normal or corrected-to-normal visual acuity and French as their mother tongue. None had experienced psychiatric or neurological
disease, or any language difficulties (e.g. dyslexia) in their past. Participants were paid for their participation (30 CHF per hour).
Experiments were performed with the informed and overt consent of each participant.
Three participants were excluded after the training session and before any EEG recording, due to low detection rates in at least one
condition. In addition, the recordings of two participants had bad impedances or muscular artefacts during the EEG acquisition and were
also excluded from further analysis. The analyses for study 1 were thus conducted with thirteen participants.
Stimuli and procedure
122
A lexical decision task was used in a simultaneous bilateral presentation. This is a revised version of the task previously investigated in
our laboratory (Ortigue et al. 2004; Mohr et al. 2005). On the screen, at 1 meter from the participant, two strings of letters were displayed
at the same time on each side of a central fixation cross, in the left visual field (LVF) and in the right visual field (RVF). Letter strings
represented either a real French word or a pronounceable pseudo-word. Two kinds of words were used: neutral and emotional words.
This resulted in five possible stimulus conditions: a neutral word in the LVF, a neutral word in the RVF, an emotional word in the LVF,
an emotional word in the RVF (all paired with a pseudo-word in the opposite side) or, two pseudo-words. All letter strings were
constituted of four black letters displayed in upper case (1.1cm high and 4.2 cm length) on a white background. The letter strings were
presented from 2.00° to 4.85° of visual angle from the fixation point on each side. Thus the two letter strings were simultaneously
displayed in distinct horizontal visual hemifields. The presentation time was 13ms. During the experiment, the participants were asked to
respond during the 2 seconds following the stimulus if they had seen a word or not. If yes, they had to press a button with the hand on the
side on which they detected the word. No response was required if they did not detect a word on either side. The inter stimulus interval
was randomly jittered between 2300 ms and 2600 ms. A sample trial is shown in figure 1.
Four letter substantives with highest frequencies (occurrences by million determined from a written database of text book material and an
oral database of movie subtitle material) were taken from the freely accessible and well controlled database for the French language
“Lexique 3.55” (New et al. 2001; see also http://www.lexique.org). Independent participants (n=12) rated these words for fami liarity,
concreteness, imageability, emotional neutrality and emotionality on a 1 to 9 Likert-scale. Only four letter words with high familiarity,
low concreteness and low imageability were kept. The 16 words with highest neutral and lowest emotional ratings were chosen for the
neutral category (e.g. the French word “voie”) while the reverse – i.e. highest emotional, lowest neutral ratings – were included in the
emotional category (e.g. the word “peur”). The selected words were matched with pronounceable pseudo-words with identical patterns of
consonants and vowels with four letters. Pseudo-words with spelling close to familiar foreign language words (commonly English and
German) were excluded to avoid any interaction in potential multilingual participants. T-tests on the selected words showed that neutral
and emotional words did not differ in terms of frequency (respective averages 92.25 vs. 95.44; with a t = -0.73 and p = 0.89), familiarity
(7.40 vs. 7.80; t = -1.88 and p = 0.07), concreteness (3.7 vs. 3.28; t = 0.85 and p = 0.40) or imageability (4.63 vs. 3.94; t = -1.24 and p =
0.23).
The task was divided into 8 blocks (4 minutes each) of 96 trials fully randomised for each block and for each participant. The total
number of repetitions was 128 for neutral words in the LVF, 128 for neutral words in the RVF, 128 for emotional words in the LVF, 128
for emotional words in the RVF, and 256 for two pseudo-words in both hemifields.
To avoid learning effects during the EEG recording due to the repeated presentation of the words, the subject practised the task in four
blocks the day before the EEG recording. Before each EEG session, the whole set of words were presented individually in a central
position below the fixation point and the participants performed a simple lexical decision task without EEG recordings. This learning
session was performed to further avoid learning effects during task performance.
Behavioural data
Average accuracies and median reaction times (RT) were calculated for each condition in each subject. These behavioural data were
plotted with their standard errors in bar-graphs (see Fig. 2). The statistical analysis of the behavioural data was performed to replicate
previous studies and was thus based on the prediction that words are better detected when presented in the RVF than to the LVF (Young
et al. 1980; Graves et al. 1981; Iacoboni and Zaidel 1996; Babkoff et al. 1997), and that emotional words are better detected than neutral
words (Landis, 2006; Schacht and Sommer 2009; Kissler and Koessler 2011; Skrandies, 2013). Missed events and false alarms were also
computed in order to detect potential outlier subjects (not represented).
123
Fig. 1. Schematic chronology in milliseconds (ms) of a trial for the presentation of one pair of letter strings in the lexical decision task in study 1 (EEG)
and study 2 (TMS).
EEG data acquisition & pre-analysis
High density EEG data were collected during 8 blocks of the task. The system used for acquisition was a 256-channel (AgCl carbon-fiber
coated electrodes) HydroCel Geodesic Sensor Net in association with Net Station EEG acquisition software (Electrical Geodesics, Inc.,
Eugene, OR, USA). Recordings were referenced to the vertex (Cz) and sampled with a 16-bit analog-to-digital converter at 1000Hz with
a DC removal. All electrode impedances were kept below 30kΩ.
Off-line analyses of the EEG were conducted using the “Cartool” software (D. Brunet, Functional Brain Mapping Laboratory and CIBM,
Geneva, Switzerland; http://sites.google.com/site/fbmlab/cartool). The signals of peripheral electrodes (on the cheeks and the nape) were
omitted from the 256 channels and the remaining 204 channels were submitted to further analysis. The EEG of these remaining channels
was bandpass filtered from 0.16 to 40 Hz and converted to the average reference. The continuous EEG was visually inspected and only
epochs without eye blinks or muscular artefacts and with correct responses were taken into account. The number of averaged epochs was
counterbalanced between the four conditions for each participant – i.e. we randomly picked a number of epochs determined by the
condition with the lowest number of accepted epochs for each individual (average number of accepted epochs: 57.2 ± 6.4 standard error).
Therefore, for a given subject, the different conditions included comparable signal to noise ratios. The remaining epochs were averaged
by condition from 0 ms to 300 ms after the onset of the stimulus to reveal the Event Related Potential (ERP) for each participant. For
each individual, artefactual electrode signals were interpolated using a spherical spline interpolation (Perrin et al. 1987). All individual
ERP’s were realigned on the peak of the Global Field Power of the P100 component determined once for each subject to reduce the interindividual variability of the time for visual information transfer from the retina to primary visual cortex (Woody et al. 1967; Morand et
al. 2000). Realignment has been - 27 up to + 30 ms from the original onset of the stimulus. As visual information does not reach the
primary visual cortex before 50 ms (Alexander and Wright 2006; Thorpe 2009), this realignment on the first milliseconds does not
impede the analysis of cognitive cortical processes, but reduces their temporal smearing. Finally ERP’s were downsampled to 500 Hz
before statistical analysis.
ERP’s Topographic ANOVA
124
Since the aim of this first experiment was to determine when and where in the brain the different stimuli were processed to guide the
subsequent TMS study, the analysis of the high-density ERPs was entirely based on topographical analysis of the scalp potential fields
and on electrical source imaging. In contrast to the conventional waveform analysis, the topographic analysis gives non-ambiguous and
reference-free information about changes of the underlying generators in time and between conditions (Michel et al. 2009; Michel and
Murray 2011; Murray et al. 2008).
The first step of the analysis consisted in a topographic analysis of variance (T-ANOVA) based on paired permutation tests with entire
topographies as dependent variables (Murray et al. 2008; Koenig et al., 2012). This analysis was performed with the freely available
software RAGU (Randomization Graphical User interface; Koenig et al. 2011), which allows for multivariate T-ANOVAs. The program
calculates the Global Field Power of the normalized difference maps between conditions for each subject, also called Global
Dissimilarity (Lehmann and Skrandies 1980; Murray and Michel 2008). The Dissimilarity is calculated repeatedly after random shuffling
the condition assignments of the maps across subjects. This permutation statistic enables to compare the real distribution of the Map
Dissimilarity against the random distribution. The significance of the effect is then determined by the percent of randomly obtained
values of the Dissimilarity that are larger than the real value. These T-ANOVAs were performed for each individual time point from 50
to 250 ms post-stimulus in a two by two within-subject factorial design (p-value < 0.05; 5000 repetitions); for words detected in the LVF
vs. in the RVF (Laterality) and for detected neutral words vs. emotional words (Emotionality).
Electrical source imaging (ESI)
In order to estimate the location of current generators in the brain, the LAURA (Local AUto Regressive Average) algorithm, a distributed
linear inverse solution comprising biophysical laws as constraints (Grave de Peralta Menendez et al. 2001; Grave de Peralta Menendez et
al. 2004; and Michel et al. 2004 for review), was used. Source estimations were computed from ERPs of each participant for each
condition according to the factors Emotionality and Laterality using the Cartool software in a space of 3005 solution points restricted to
the gray matter of the MNI brain (Spinelli et al. 2000; Brunet et al. 2011). For each participant and each condition, a normalisation of the
estimated activity values was performed by z-score transformation of the activity at each solution point for a given time point. The
transformation in z-scores homogenised inter-individual contrasts by removing the background noise and reducing extreme values. Using
these z-scores, as dependent variables, time point by time point ANOVAs were computed for each solution point (p < 0.01 for at least 5
consecutive time frames corresponding to 10 ms and with a size restriction criterion of 5 neighboured solution points) with the STEN
toolbox (STEN toolbox developed by J.-F. Knebel, Functional Electrical Neuroimaging Laboratory, Lausanne, Switzerland.
www.unil.ch/fenl/Sten). Laterality and Emotionality of the words were taken as within-subject factors. Post-hoc analyses, consisting in ttests on each solution point for each time frame, were computed to define the direction of the significant differences established by the
ANOVA.
Study 2 (TMS)
Participants
Thirteen healthy volunteers participated in this study approved by the Ethical Committee of the University Hospital of Geneva,
Switzerland. These participants did not take part in study 1 (EEG). They were young adult males, right-handed, aged from 21 to 33 years
old (mean age = 24.54; SD = 4.14). All had normal or corrected-to-normal visual acuity, French as their mother tongue and reported no
language impairments (e.g. dyslexia). None had experienced psychiatric or neurological disease or unexplained loss of consciousness,
and none reported having experienced any epileptic episodes in their past or having a history of epileptic events in their family. The
experiment was performed with the informed and overt consent of each participant. Participants were paid for their participation (30 CHF
per hour) and were free to withdraw from the study at any time.
125
Results of three participants were excluded from analysis due to outlier detection levels or a high level of false alarms. Two participants
had high levels of false-alarms (wrong detection in pseudo-words condition) in all three conditions of TMS indicating that they had
difficulties discriminating words from pseudo-words or were not focused on the task. In both cases the false alarm levels were considered
as outliers according to the mean plus or minus two standard deviation criterion. One other participant had extremely high levels of
average detection which also met the criteria for outlier exclusion – i.e. 2 standard deviations above the group mean. The analyses for the
study 2 (TMS) were thus conducted on a sample of ten participants.
Stimuli and procedure
The same simultaneous bilateral lexical decision task as in study 1 (EEG) was used with the same stimuli presented in the same manner
and with the same instructions. The only difference with the task from study 1 was that the inter stimulus interval was randomly jittered
from 3340 to 3640 ms – i.e. an extended period to avoid cumulative effects of TMS between trials. Responses of the participants were
collected during 2500 ms (Fig. 1).
The task was divided into blocks of 48 trials, freely randomised for each block and for each participant. 8 neutral words in the LVF, 8
neutral words in the RVF, 8 emotional words in the LVF, 8 emotional words in the RVF, and 16 pairs of pseudo-words were presented
per block. The day before the TMS experiment, participants were asked to perform four blocks as behavioural measurement and training
session without any TMS involvement. Then, on the following day, they had to execute three blocks of the task while TMS was released
over each site of stimulation (3 blocks x 3 sites of stimulation). In addition, each word was first displayed as single word in a central
position below the fixation point and subjects performed a simple word detection task. This was done to avoid learning effects during the
task due to the repeated presentation of a word.
Transcranial Magnetic Stimulation parameters
The transcranial magnetic stimulation parameters were in accordance with the current safety guidelines of TMS use (Rossi et al. 2009). A
Magstim Rapid2 with a double 70 mm figure-of-eight coil (The Magstim Company Ltd, Whitland, Wales, UK) was used for magnetic
stimulation. TMS was performed as an event-related paradigm to interact transiently with cortical areas of interest during the task. A
train of three bipolar pulses at 20Hz was released 50 ms after each onset of the stimulus – i.e. the pulses were delivered 50, 100 and 150
ms post-stimulus (Fig. 1). Hence, the TMS pulses are delivered after the display of the letter strings, thus avoiding alerting effect s on
perception due to sound or skin sensations induced by TMS. The strength of stimulation was set individually to 90% of the resting motor
threshold (RMT) intensity – i.e. the intensity, which generates a finger movement for more than half of the motor cortex stimulations.
Reduced intensity is known to produce a more focal effect (Wagner et al. 2009). In addition, a moderate intensity of stimulation limits
participant discomfort and muscle contraction.
Both, the time period and the sites of stimulation were based on the results of the EEG study (see results section 3.2.1). On a different
day prior to the TMS sessions, the participants underwent a T1-weighted structural MRI scan. The target sites for TMS stimulation were
determined based on individual structural MRI on the anterior bend of the angular gyrus – Brodmann area 39, in the middle of the
temporoparietal junctions (TPJ) of both hemispheres. The average Talairach coordinates of the actual targets across participants were 58.9; -54; 25.8 for the left angular gyrus (LAG) and 58.2; -50.7; 27.8 for the right angular gyrus (RAG). The vertex (0; -25; 68) was used
as an active control site. A frameless neuronavigation system combining the neuronavigation module of BrainVoyagerQx
(Braininnovation, Maastricht, The Netherlands) with the ultrasound CMS20 measuring system for navigation (Zebris GmbH, Tübingen,
Germany) and based on individual structural MRI was used to place and to maintain the coil over the considered site for every single task
block. The three different locations of the stimulation were alternated following a sequence of nine blocks in a single-blind design;
sequenced as a/b/c/a/b/c/a/b/c. The stimulation sites were randomly attributed to a, b or c between participants. For the two lateralised
126
positions, the direction of the currents flowing through the coil at the junction of the two wings was oriented from superior-medial to
inferior-lateral. In the vertex condition, the currents were oriented forward.
Analysis of behavioural data
Average accuracy and median RT of correct detections were calculated for each stimulus condition and each TMS condition in each
subject. These behavioural data were computed for conditions with real words using a three-way ANOVA with the within-subject
factors: TMS – i.e. site of stimulation: LAG vs. RAG vs. vertex, Laterality – i.e. side of presentation: LVF vs. RVF, and Emotionality –
i.e. valence of word: neutral vs. emotional. Missed events and false alarms were also computed in order to detect potential outlier
subjects (not represented).
The accuracy and median RT for the LAG and RAG stimulations were individually normalised by the behavioural variables obtained for
the control vertex stimulation as follows:
Vertex Normalized Index = (variable LAG or RAG - variable Vertex) / (variable LAG or RAG + variable Vertex)
The variables thus obtained were vertex normalized indices (VNI). In the case of accuracy, a positive VNI reflects an increase in
detection performance caused by TMS over LAG or RAG compared to accuracy obtained after vertex stimulation. Concerning RT, a
positive VNI indicates a slowing of the response speed. This normalization reduces unspecific side effects of TMS that could be due to
peripheral sensations as well as the alerting properties of the loud click sound of the TMS stimulation (see also Romei et al. 2011;
Rochas et al. 2013). Note that this index is unit-free.
The VNI for accuracy and median RT were computed using a three-way ANOVA (2 x 2 x 2) with the within-subject factors: TMS – i.e.
site of the TMS: LAG vs. RAG, Emotionality – i.e. valence of the word: neutral vs. emotional, and Laterality – i.e. side of presentation:
LVF vs. RVF.
RESULTS
Study 1 (EEG)
Behavioural results
Figure 2 indicates the mean and standard error of the accuracy and median RT in the EEG study. In view of our analysis, the results for
both accuracy and median RT providing two tests for each effect, a p-value < 0.025 critical level was chosen to maintain an overall pvalue < 0.05 error rate. Greenhouse-Geisser corrections were employed for all effects. As can be seen in Figure 2, both the expected RVF
advantage (accuracy: F(1,12) = 12.82; p-value = 0.004 and RT: F(1,12) = 6.28; p-value = 0.028 n.s.) and the expected emotional word
advantage (accuracy: F (1,
12)
= 0.49; p-value = 0.50 n.s. and RT: F(1,12) = 17.09; p-value = 0.001) were present. An interaction of
emotional word advantage and visual field was observed (accuracy: F (1, 12) = 31.67; p-value < 0.001 and RT: F(1, 12) = 1.46; p-value = 0.25
n.s.).
127
Fig. 2. Averages and standard errors of the individual accuracies (a) and median RT in ms (b) for the four stimulus conditions with words (left or right
visual field neutral word, and left or right visual field emotional word) in study 1 (EEG) (n = 13).
ERP’s Topographic ANOVA results
The time-point by time-point T-ANOVA revealed significant differences for the factor Emotionality from 116 to 134 ms after the
stimulus onset (Fig. 3A).The difference map indicated increased positivity over anterior electrodes for neutral words and increased
positivity for emotional words over posterior electrodes for this time window. Based on the global field power of the grand-mean ERPs
(not represented), this time period falls in the transition between the P100 component and the subsequent N170-like component.
The comparison for the factor Laterality showed differences in a large time window, starting just 52 ms after stimulus presentation and
covering the entire P100 component – from 52 to 140 ms (Fig. 3B). In this period, the difference map between the LVF and RVF
conditions showed a clear left-right polarity inversion with relative stronger right positivity when the real word was detected in the LVF
and the pseudo-word in the RVF, and a left positivity in the reversed condition. The corresponding averaged maps during this time
period, resembling the visual P100 topography, had left and right side maxima of positive potentials that were inverted between the two
conditions, with more positivity over the right posterior electrodes for LVF words and more positivity over left posterior electrodes for
RVF words. The two conditions differed again in a second time window starting at 184 ms – from 184 to more than 250 ms. This time,
the difference map showed inverted polarities over frontal electrodes with more positivity over right frontal electrodes for words in the
LVF and more left frontal positivity for words in the RVF.
There were interactions between the two factors lasting less than 10 ms at 130 ms, then more durably from 158 to 202 ms (Fig. 3C).
Interactions signified that differences of map configurations for the Emotional factor are dependent on the difference according to the
side of presentation; in other words, the differences of processing between neutral and emotional words were dependent on the visual
field in which they were presented. The scheme of the interaction opposes any real word in the RVF and neutral words in any visual field
against words in the LVF and emotional words; this indicates that words in the RVF share similar brain states with neutral words, while
words in the LVF show a similar response to emotional words.
Critically though, the chronology of these effects suggests that map topographies differed first as a function of the side of presentation of
the word (50 ms post onset), then as a function of the emotional content of the words, and finally with an interaction of these two factors
after 150 ms. From 210 ms onwards, the only remaining differences were for the side of word presentation.
128
Fig. 3. Topographic ANOVA comparisons of data according to the emotional nature (emotionality) or the side of presentation (laterality) of words and
their interaction across time from 50 to 250 ms after the onset of the stimulus. Black solid rectangles depict time points of significance (p < 0.05), and the
topographic maps (μV) show the maps averaged for these time periods for the corresponding conditions as well as the difference maps between them (n =
13)
Electrical source imaging (ESI) results
The analysis in the inverse space focused on those time periods that revealed significant topographic differences in the analysis of the
scalp maps – i.e. around the P100 component and the N170-like component.
Considering the first period of interest – i.e. from 70 to 134 ms, the analyses of variances on the z-scored current density values revealed
differences for the main factor Laterality similarly to the scalp map analysis. These differences were found mainly in the ri ght Precuneus
– Brodmann area 7 (Fig. 4A), and somewhat weaker in the right inferior parietal lobule and right angular gyrus – Brodmann area 39, and
in the left superior and middle temporal gyri – Brodmann areas 21, 22, 41 and 42. Post-hoc analysis revealed a stronger activity in these
areas for words detected in the LVF as compared to the RVF.
Concerning the second time period of interest – i.e. from 130 to 210 ms, the analysis of variance revealed differences for both main
factors Laterality and Emotionality as well as for the interaction between the two factors. The first differences were found for the factor
Emotionality in the very beginning of the period from 132 to 156 ms, in a large cortical area from the right inferior to the superior
parietal lobule – Brodmann areas 7, 19, 39 and 40. Post-hoc analyses revealed stronger activity for emotional words than for neutral
words in these areas (Fig. 4B). The factor Laterality differed from 148 ms onwards. From 148 to 170 ms the differences were found first
in the right supramarginal gyrus – Brodmann area 40, and in the left post-central gyrus. Afterwards, from 172 to 200 ms differences were
found in an area connecting the posterior part of the right middle temporal gyrus, the right angular gyrus and the anterior part of the right
extrastriate cortex – Brodmann areas 19, 37 and 39, in the left Insula – Brodmann area 13, and in the left inferior frontal gyrus. All these
areas of divergence were found to be more active for the detection of words presented in the LVF than in the RVF, the most relevant area
being the right middle temporal gyrus (Fig. 4C). Finally the ANOVA revealed interaction effects at the end of the period from 172 to 206
ms, with differences located in the left precuneus and in the posterior part of the left middle temporal gyrus – Brodmann area 21 (not
represented). The interaction was mainly due to stronger activity for emotional words detected in the LVF than neutral words detected in
the RVF.
129
Fig. 4 Graphical representations on horizontal sections of MNI template brain of the significant differences (p < 0.01 for at least 10 ms) from ANOVA of
the z-scored current density values according to the factors Emotionality (Neu/Emo) and Laterality (LVF/RVF), weighted by the t values from post hoc
analyses and averaged over time periods. Warm colours depict differences in favour of the first group (Neutral or LVF), and cold colours depict
differences in favour of the second group (Emotional or RVF). Brain sections were represented in neurological convention (“le ft is left”). For an easier
reading of the obtained differences, graphical representations were condensed on the maximal difference averaged on time periods of interest based on the
topographical results (n = 13)
Study 2 (TMS)
EEG-based TMS target choice
The targets of the TMS and the timing of stimulation were both based on the results of the EEG study. As described in the EEG results
section above, the main differences identified by the z-score statistics of ESI were recurrently located in the right temporoparietal area
during the periods of interest from 70 to 200 ms (Fig.4). More specifically, from the beginning of significance of the Emotionality factor
at 132 ms these differences in the right hemisphere were found on a triangle delineated by the posterior part of the right mi ddle temporal
gyrus, the right superior parietal lobule and the anterior part of the right extrastriate cortex. Therefore the right temporoparietal junction
was chosen as a target area for TMS and more precisely in the angular gyrus. Interestingly, according to the averages of the z-scores of
ESI (not represented) and following the initial activation of the primary visual cortices, the left TPJ was consistently acti vated in any
condition while the right was more activated by emotional and LVF words – as suggested by the statistical analysis in ESI. Consequently
we decided to test the implication of the two hemispheres by stimulating two homologous areas in the TMS study. These bilateral
activations were present from around 70 ms. We chose to initiate the stimulation at 50 ms in order to disturb the focused area before the
reception of information – i.e. the transfer from the primary visual cortex to the ipsilateral TPJ – and to curtail the propagation to other
areas.
Behavioural data and Vertex Normalised Index
The mean and standard error of the accuracy and median RT are resumed in table 1. The volunteers did not report any difficulties
executing the task during TMS and accuracy was above chance (> 0.33). In view of our analysis, the results for both accuracy and
median RT providing two tests for each effect, a p-value < 0.025 critical level was chosen to maintain an overall p-value < 0.05 error
rate. Greenhouse-Geisser corrections were employed for all effects. As in study 1, the data demonstrated a replication of the well-known
advantage for the detection of words in the RVF compared to the LVF(accuracy: F (1, 9) = 55.26; p-value < 0.0001 and RT: F(1, 9) = 22.77;
p-value = 0.001), and for emotional words compared to neutral words (accuracy: F (1, 9) = 0.96; p-value = 0.95 n.s. and RT: F(1, 9) = 10.13;
p-value = 0.01). Neither an interaction nor an effect of the factor TMS were found.
130
The mean and standard error of VNI for accuracy and median RT are summarized in table 2. In view of our analysis, the results for both
accuracy and median RT providing two tests for each effect, a p-value < 0.025 critical level was chosen to maintain an overall p-value <
0.05 error rate. Greenhouse-Geisser corrections were employed for all effects. Results of the ANOVA with the VNI of accuracy did not
show independent effects of a factor or any interaction. The results for the ANOVA with the VNI of the RT also did not show any
independent effects of the different factors – i.e. location of TMS, Laterality (side of presentation) or Emotionality. However, there was
an interaction between the factors Laterality and Emotionality (F (1, 9) = 8.02; p = 0.02). While the limited statistical power provided by
the n = 10 sample size precludes follow-up post-hoc statistical tests of this interaction, the basis for the interaction appears from figure 5,
the detection of emotional words presented to the LVF was more slowed by TMS in comparison to the detection of neutral words, while
the reverse was true for words presented to the RVF.
Table 1 Averages and standard errors of the individual accuracies and median RT for the four stimulus conditions with words (left or right visual field
neutral word, and left or right visual field emotional word) and the three conditions of TMS (vertex, left AG, and right AG) in study 2 (TMS) (n = 10).
Table 2. Averages and standard errors of the vertex normalised index (VNI) of accuracy and median reaction times (RT) for TMS over left and right
angular gyri (AG) for the four stimuli conditions with words (left or right visual field neutral word, and left or right visual field emotional word) in the
study 2.
131
Fig. 5. Averages and standard errors of the vertex normalised index (VNI) of median reaction times (RT) merged between left and right TMS for the four
stimulus conditions with words (left or right visual field neutral word, and left or right visual field emotional word) in study 2. A positive VNI indicates
slowed RT compared with vertex stimulation (n = 10).
DISCUSSION
Performances in word detection change from one individual to another (Levy et al. 2010) and as a function of the stimuli or their
environment. Differences of the type of words and the conditions in which they are presented lead to multiple detection strat egies
reflected also by differences in performance (Riba et al. 2010). The location of the word in the visual field will define by which cerebral
hemisphere the visual information is initially treated. The brief simultaneous bilateral presentation used in our experiments allows a
single letter string to be initially projected to its contralateral target hemisphere as segregated information since it excludes reflexive
saccades during these very short presentation times. While this dissociation at the entry stage might appear unnatural, the distinction
between hemispheres for word detection constitutes an interest for the understanding of the functional lateralisation and the balance
between the two hemispheres in the case of challenging situations in the healthy brain or following brain injuries such as stroke. The
other factor tested in the present study was the emotionality of the words. Emotionality is known to lead to an advantage in detection
compared to neutral conditions and to interfere with the processing of word reading (Schacht and Sommer 2009; Kissler and Koessler
2011; Skrandies, 2013). Since a lexical analysis would be assumed to start before a word can be characterized as emotional, this early
differentiation for emotional word detection is surprising. The current study confirmed and refined the previous findings on the
specificity of processing of emotional words and the role of the right hemisphere in lexical decision (Ortigue et al. 2004) by further
analyses with new algorithms and extended them by showing that TMS during these early latencies over the temporoparietal areas
selectively influences emotional word processing. In addition it revealed a cooperative engagement of these homologous areas. Our
results on specific early bilateral processing of written words will be integrated in a spatiotemporal model of cerebral mechanisms.
Early bilateral processing of words
The behavioural data show that the participants recognized the words – accuracies were above chance in all conditions for the two studies
(Fig. 2 and Tab. 1), denoting non-subliminal phenomena in spite of the rapid presentation of the stimuli. Moreover behavioural results in
the two studies show that our paradigm was in accordance with the well-known advantage for the detection of words presented to the
132
RVF in the right-handed male subject (Young et al. 1980; Graves et al. 1981; Iacoboni and Zaidel 1996; Babkoff et al. 1997) and the
advantage for the detection of emotional words compared to neutral words (Landis 2006). For any trial, each visual hemifield, and
therefore each opposite hemisphere, receives a four letter string flashed exactly in the same fashion. The only remaining difference is the
significance of these letters – i.e. they are a pseudo-word or a real word, neutral or emotional. Words are detected better and faster when
they are presented in the RVF compared to the LVF, and they are also detected faster when they are emotional. These behavioural results
reflect the electrophysiological signals. In spite of apparent similar global topographic maps and electric source images in all conditions –
a certain similarity is to be expected when physical parameters are identical between conditions – the inter-condition contrasts stand out
with obvious differences. The cerebral processing is already different at early stages, during the P100 component, according to the side of
presentation of a word and according to its emotional content. These results agree with a rapid access to the lexical dimension after
stimulus onset reported in previous literature (Scott et al. 2009; Hauk et al. 2009; Segalowitz and Zheng 2009; Skrandies, 2013). These
differences occur, however, much earlier than the classically reported time for letter string processing around 150 ms (Grainger and
Holcomb 2009; Pylkkanen and Marantz 2003). Moreover, the divergences found in our bilateral lexical decision studies in EEG and
TMS argue for a specific detection at early stages in each hemisphere. These early latencies argue for an extremely fast start of the
discrimination process before the crossing of the information to the opposite hemisphere is likely. Similar early differences in the ERPs
between 80 and 150 ms depending on the connotative meaning of visually presented words have been described by Skrandies in adults as
well as in children (Skrandies, 2013). The fact that in our study the words are very short and familiar plays a probable role in the
swiftness of processing. Dissimilarities between the detection of a word in the LVF vs. in the RVF shown by topographic analysis are
mainly reflected as differences in left-right lateralisation of the polarities (Fig. 3B) during the two time windows of interest – i.e. P100
and N170. More than a simple left-right inversion of the pattern of activations in the brain, the ANOVA on the z-scored current density
values explains this persistent dissimilarity in topography with additional generators initially located in the right precuneus with a
supplementary activation for words detected in the LVF (Fig. 4A), in addition to the standard network activated by the detection of
words in the RVF. While the P100 mainly corresponds to the processing of visual information in the visual cortex, these results argue for
differences that are due to concomitant pre-processing of word material in areas beyond the primary cortex. The precuneus has functional
connections with 1) the visual cortex from its posterior part along the parieto-occipital fissure and 2) the inferior parietal lobule and the
superior temporal sulcus from its central part near the precuneal sulcus (Margulies et al. 2009; Zhanga and Li 2012). The expected
activation of the visual cortex by words flashed in both hemifields is followed by activation of the inferior parietal lobule and the
superior temporal sulcus for further steps of word processing. Indeed, the divergences are then located in the areas adjacent to the
occipital lobe in particular with the right TPJ around 130 ms and the right middle temporal gyrus at 170 ms (Fig. 4B and C). Again, the
results indicate supplementary activations in the right hemisphere respectively for the detection of emotional words and words presented
in the LVF.
A delayed but superimposed processing of the emotionality
The assumption of a lexical detection process in the right hemisphere for words presented to the LVF would not have been vali dated by
the isolated analysis of the Laterality factor in the EEG study. Indeed the observed difference between the two conditions, LVF vs. RVF,
at a given time could be solely due to a detection of a word in the RVF by the left hemisphere vs. a non detection in the other. However,
the results from the ANOVA on the z-scored current density values indicate functional activations of the right hemisphere. Moreover the
early divergences in the processing of emotionality (Fig. 3A and 3B) and the interactions involving the factors Emotionality and
Laterality point to an actual early detection of words also in the LVF and different detections in both hemispheres. The TMS results
argue in favour of a cooperative network between both hemispheres that is balanced according to the two factors. Specific effects for
emotional words in the LVF point to a semantic treatment.
Compared to the Laterality factor, the differences due to the factor Emotionality in the EEG study are more focused in time a nd appear a
little later at 116 ms in the topography analyses (Fig. 3A) – orthogonally to what was observed in the results for the factor Laterality.
133
Results from the ANOVA on z-scores of current density values show brief but strongly significant differences for the factor Emotionality
between the P100 and the N170-like components, from 132 ms onwards, with a maximum of increased activity for the detection of
emotional words compared to neutral words in the right TPJ (Fig. 4B). Hence the emotional differentiation might thus occur downstream
of the primary visual cortices. Interestingly the TPJ is connected with the precuneus and also with the primary visual cortex (Richardson
et al. 2011; Zhang and Li 2012).
Additionally, the slight emotional advantage in detection in the RVF in study 1 (Fig. 2) could be due to the observed bilateral implication
of the TPJs. Indeed, RVF emotional words could be processed in the left hemisphere as they come from the RVF and by an extra
processor in the right hemisphere for their Emotionality (Fig. 4B). Therefore the two sites involved in RVF emotional word processing
might lead to a redundant lexical treatment and reduce missed events while only the right side is involved in the processing of emotional
words from the LVF in the conditions of our experiment – i.e. with short, abstract and very familiar words. Interestingly, the absence of a
significant advantage in RT for these emotional words coming from the RVF (Tab. 1) is in accordance with a model of double processing
in the left and the right hemispheres. The bilateral processing of these words would be better but not necessarily faster due to multiple
exchanges between brain areas.
Functional interaction between the conditions
Obviously the brain computes words in the LVF or the RVF as well as neutral or emotional words in a different way. Nevertheless, the
interactions found in the EEG and TMS results during the first 250 ms of processing suggest that word detection varies as a function of
Emotionality and Laterality in a dependent and similar way and indicate a difference but not a clear distinction. More precisely, the
processing of words in the LVF is related to the processing of emotional words while the detection of neutral words and of words
presented in the RVF would also converge to similar brain states. A possible reason could be that the treatments of these two
characteristics interact in the processing of word detection because they share similar structures or networks – as their disruption by
short-burst TMS tends to assert. In the case of regular word detection, some unavoidable relays may be used. The interactions would be
generated by interdependent collisions (interactions) in these common neural substrates. As developed hereafter, the TPJ regi ons might
be at the centre of a possible common network. As TMS is released from 50 to 150 ms, the physiological functions of the target area and
possibly its connected areas are degraded during this time period and the following decades of milliseconds (Valero-Cabré et al. 2007).
TMS applied over the left or right angular gyrus does not disturb the key tendencies of the behaviour of the participants. The detection is
still better for words presented in the RVF compared to LVF and emotional words are detected faster than neutral words regardless of the
TMS conditions (Tab. 1). However compared to what is observed for the vertex stimulation, TMS modifies the reaction times when
applied over the left or right angular gyrus according to the factors Laterality and Emotionality (Tab. 2). Nevertheless as the slowing is
not the same for all categories of stimuli, the implication of the angular gyri is stimulus-dependent. Indeed, the behavioural perturbation
mediated by the disruption of the angular gyri in the first 200 ms is an interaction indicating that emotional word detection in the LVF is
more impaired by TMS in comparison to the detection of neutral words. This is supported by the difference between neutral and
emotional word detection for the LVF while the detection of words in the RVF shows a weaker difference (Fig. 5). In this sense, the
detection of words in the LVF would have a more weight in the interaction found in the ANOVA of VNI for median RT. The TMS
evidence indicates that the processing of emotional words in the LVF can be impaired by acting on bi-hemispheric targets of the TPJs,
namely the angular gyri. Both TPJs might have cooperative functional roles in word detection as already suggested by differences found
in both hemispheres in the ESI analyses. Notably, previous studies using non-repetitive TMS showed rapid contralateral propagations of
TMS-evoked activations to the homologous area in the motor cortex (Ilmoniemi et al. 1997; Parks et al. 2011; Casali et al. 2013;
Ragazzoni et al. 2013), the premotor cortex (Massimini et al. 2005) as well as the visual cortex (Ilmoniemi, Virtanen, et al. 1997).
Although all these studies used single pulse stimulation protocols, a transcallosal propagation with our triple-pulse stimulation that would
lead to the observed absence of distinction between left and right TMS cannot be fully ruled out. The right side that is specialised for
emotional processing could thus also be disturbed by the stimulation of the left side. In this case, the effect of the TMS of the left TPJ
134
could be due to the impairment of the activations found in the ESI analyses in the right TPJ. While differences between TMS over the
left or the right TPJs did not reach statistical significance, the stronger slowing of the detection of emotional words from the LVF due to
the disruption of the right TPJ (highest VNI for median RT in Tab. 2) could result from a main contribution of this structure in the
network activated by the detection of emotional words and words in the LVF as revealed concordantly in the ESI analyses (Fig. 4). This
last point would indeed argue in favour of an effect of TMS over the left TPJ mediated by the propagation of disturbance to the right.
Further combined TMS-EEG studies during task execution are needed to fully clarify the mechanisms underlying the observed effects.
A left-right balance
The results from the EEG and TMS experiments argue for a variable implication of both hemispheres in the detection of written words
balanced by the conditional involvement of the right hemisphere according to different conditions. Indeed the main areas of divergence
in the EEG analysis are found in the right hemisphere – in or close to the right TPJ – across both time periods of interest. The TMS
results confirm the conditional implication of the right TPJ and suggest a possible cooperation with the left homologous TPJ. Thus the
right hemisphere would be more active for the detection of emotional words and words presented in the LVF. The areas thereby involved
in the right hemisphere would be part of the process in addition to the regular network which is located in the left hemisphere.
Conversely, the balance would be all the more in favour of the left hemisphere for neutral words and words coming from the RVF, with a
reduced intervention of the right hemisphere during early stages of word processing.
The recruitment of the right hemisphere observed for emotional words is in accordance with an important part of the literature
demonstrating the role of the right hemisphere in emotional functions in general (from Gainotti 1972 and 2005 with the “right
hemisphere hypothesis” to the adapted version of the “valence hypothesis” of Davidson 2001for the comprehension of emotional
information) and more interestingly in the processing of emotional words (Landis 2006). Numerous studies were conducted in aphasic
patients showing an advantage for emotional word processing with a lesion of the left hemisphere (Landis et al. 1982; Landis et al. 1983;
Reuterskiöld 1991) or specific emotion processing impairments with a right hemispheric lesion (Borod et al. 1992; Lalande et al. 1992;
Cicero et al. 1999). In addition to the left lateralised language network, the processing of word-related emotionality is also found in the
right hemisphere in healthy participants – e.g. Ortigue et al. 2004; Frühholz et al. 2011; Ponz et al. 2013. Our results argue for
supplementary processing performed by the right hemisphere at very early stages. Moreover this lexical emotion processing in the right
hemisphere is dependent on the connexion with the left hemisphere. This interhemispheric cooperation could explain paradoxical left
sided processing of emotionality in words observed in some studies (Kissler et al. 2007; Herbert et al.2008).
The role of the TPJ
The TPJs could be a part of the networks as either a pathway, or a processor or both. The fact that the TMS-induced disturbance also
includes Emotionality supports an advanced treatment of words in both TPJs or at least in the related upstream areas, and would disagree
with a simple pathway of the information that does not involve word processing. The left TPJ is known to be implicated in language
comprehension (for review Price 2012). Delimited parts of the TPJ have shown their contribution for specific aspects of visual word
processing, such as the supramarginal gyrus in the phonologic dimension of visual word recognition (Stoeckel et al. 2009), the angular
gyrus in response to semantic vs. phonologic aspects (Binder et al. 1999) or in multimodal processing (Bonner et al. 2013) while the
posterior superior temporal gyrus is rather implicated in word sound processing (Buchsbaum et al. 2001). A bilateral involvement is
depicted in the STG notably for speech intonation processing and the right temporal lobe is also more engaged in speech proce ssing
during sentence and context comprehension (Vigneau et al. 2011; Diaz and Hogstrom 2011). Bilateral activation of the angular gyri and
superior temporal gyri is found in word reading compared to non-words (Kuchinke et al. 2005) or angular gyri alone in high frequency
word reading (Graves et al. 2010). In the present work, the two TPJs engage a real cooperation which is conditioned by the stimulus
135
features to achieve word reading. The right hemisphere is involved for LVF words and emotional words implying a bilateral processing
with an interaction between the two hemispheres as shown by interaction in the ESI results and suggested by TMS. Accordingly,
Vigneau et al. (2011) also found an implication of the right hemisphere in an inter-hemispheric manner in association with left
homologous areas.
Spatiotemporal model
Based on our studies, a spatiotemporal model may be proposed that would explain the path of information processing in four steps. The
information from each visual hemifield reaches each primary visual cortex separately 50 ms after the stimulus display as revealed by the
averaged z-scores of ESI. From here, divergences are observed in the processing of word detection with maximal divergences around 100
ms in the right precuneus and slightly in the right inferior parietal lobules according to Laterality. Average activities in z-scores of ESI
are already found in both TPJs before 100 ms. As a first step, a precursor of the lexical decision involving the classification of a letter
string as a word – or not – is elicited in the receptor hemisphere from 50 to 130 ms. Activity is concentrated around the superior temporal
sulci with diverging processing in terms of Laterality in the right supramarginal gyrus from 160 ms onwards and then in the superior
temporal gyrus – i.e. higher activity is observed for words detected in the LVF. In the meantime, a precursor of a semantic analysis of the
letter strings has to be initiated to enable further computation of their emotional characteristics. The emotional detection appears to start
with first differentiations between neutral and emotional words at 130 ms. In this second step, while the TPJs are already engaged in
lateralized processing, the influence of Emotionality emerges concomitantly from 110 to 150 ms. The superior part of the right TPJ
shows a specific treatment of emotionality from 132 to 156 ms with a higher activity found in the inferior parietal lobule for emotional
words. At this point the processing seems to be specific in each hemisphere while also integrating information from interhemispheric
communication. Indeed, the activation suggests an inter-hemispheric cross-over of information – i.e. for emotional words from the RVF.
From there the interaction between the two factors appears at 130 ms and then from 160 to 210 ms. The interaction could indicate a
conflict between the processing of different factors in common areas of the two hemispheres from precuneus to temporal lobes via
inferior parietal lobules. The interaction found in the TMS study tends to confirm the scenario of this bilateral communication with the
contribution of the left and right TPJ. From 210 ms onwards, the differences were found for Laterality only in the right TPJ and in the
primary visual cortex, followed by activations in the primary visual cortex and the left middle temporal gyrus (not related in the results
section). This could imply further computations that are related to the motor response required by the task, however, the ESI would rather
suggest a feedback process with activations and differences located in primary visual areas which might enable integration for further
semantic processing. From this point onwards the processing would eventually continue in the left hemisphere.
CONCLUSIONS
In conclusion, our results show that lexical detection processing starts very early in both hemispheres. A bilateral network is required for
the detection of words according to the side of presentation and their emotional content. Moreover analysis of slowed reaction times in
the TMS experiment indicates that interference with both left and right temporoparietal junctions results in impaired processing of
emotional words that were presented to the left visual field. Our research confirms the hypothesis of an early cooperation between the
two hemispheres conditioned by word features. The complementarity between the left and the right side of the brain together with their
interaction in word processing constitute a critical point in the understanding of reading in healthy subjects but also in the case of
recovery from injury.
ACKNOWLEDGEMENTS
136
This research was supported by the Swiss National Science Foundation: grant No. 320030_132967 to Theodor Landis and grant No.
310030_132952 to Christoph M. Michel. The TMS equipment and the Cartool software are supported by the Center for Biomedical
Imaging (CIBM), Geneva and Lausanne, Switzerland. The Cartool software was developed by Denis Brunet from the Center for
Biomedical Imaging (CIBM), Geneva and Lausanne, Switzerland. http://sites.google.com/site/fbmlab/cartool
The STEN toolbox (http://www.unil.ch/fenl/Sten) was programmed by Jean-François Knebel, from the Functional Electrical
Neuroimaging Laboratory, Lausanne, Switzerland, and is supported by the Center for Biomedical Imaging (CIBM) of Geneva and
Lausanne and by the National Center of Competence in Research project “SYNAPSY – The Synaptic Bases of Mental Disease”; project
no. 51AU40_125759.
The authors would like to thank Alexis Hervais-Adelman who assisted us with the analysis of selecting a balanced number of epochs
between conditions in study 1 (EEG).
Special thanks go to Gregor Thut who helped to initiate this project and provided precious input and expertise during all stages of the
project.
REFERENCES
AlexanderDM, Wright JJ. (2006) The maximum range and timing of excitatory contextual modulation in monkey primary visual cortex.
Vis Neurosci, 23:721-728.
Babkoff H, Faust M, Lavidor M. (1997) Lexical decision, visual hemifield and angle of orientation. Neuropsychologia, 35(4):487-95.
Binder JR, Frost JA, Hammeke TA, Bellgowan PSF, Rao SM, Cox RW. (1999) Conceptual Processing during the Conscious Resting
State : A Functional MRI Study. J Cogn Neurosci, 11(1):80-93.
Bonner MF, PeelleJE, Cook PA, Grossman M. (2013)Heteromodal conceptual processing in the angular gyrus. NeuroImage, 71:175-186.
Borod JC, Andelman F, Obler LK, Tweedy JR,Welkowitz J. (1992) Right hemisphere specialization for the identification of emotional
words and sentences: evidence from stroke patients. Neuropsychologia, 30:827-844.
Broca P. (1861) Remarques sur le siège de la faculté du langage articulé, suivies d’une observation d’aphémie (perte de la
parole)Bulletins de la Société Anatomique de Paris, 6:330-357.
Brunet D, Murray MM, Michel CM. (2011) Spatiotemporal analysis of multichannel EEG: CARTOOL. ComputIntellNeurosci,
2011:813870.
Buchsbaum BR, Hickok G, Humphries C. (2001)Role of left posterior superior temporal gyrus in phonological processing for speech
perception and production. CognSci, 25(5):663-678.
Casali AG, Gosseries O, Rosanova M, Boly M, Sarasso S, Casali KR, Casarotto S, Bruno M-A, Laureys S, Tononi G, Massimini M.
(2013) A theoretically based index of consciousness independent of sensory processing and behavior. Sci Transl Med, 5(198):
198ra105.
Cicero BA, BorodJC, SantschiC, ErhanHM, Obler LK, Agosti RM, Welkowitz J, Grunwald IS. (1999) Emotional versus nonemotional
lexical perception in patients with right and left brain damage. Neuropsychiatry Neuropsychol Behav Neurol, 12:255-264.
137
Davidson RJ. (2001) Prefrontal and amygdala contributions to emotion and affective style. Handbook of neuropsychology, Emotional
behaviour and its disorders. 5(2nd ed.): 111–12. Amsterdam (NL). Elsevier, G Gainotti.
Diaz MT, Hogstrom LJ. (2011) The influence of context on hemispheric recruitment during metaphor processing. J CognNeurosci,
23(11):3586-3597.
Frühholz S, Jellinghaus A, Herrmann M. (2011) Time course of implicit processing and explicit processing of emotional faces and
emotional words. BiolPsychol, 87(2):265-74.
Gainotti G. (1972) Emotional behavior and hemispheric side of the lesion. Cortex, 8(1):41-55.
Gainotti G. (2005) Emotions, unconscious processes and the right hemisphere. Neuro-Psychoanalysis, 7:71–81.
Gazzaniga MS, SperryRW. (1967) Language after section of the cerebral commissures.Brain, 90:131-148.
Gazzaniga MS,HillyardSA. (1971) Language and speech capacity of the right hemisphere.Neuropsychologia, 9:273-280.
Gazzaniga MS. (2005) Forty-five years of split-brain research and still going strong. Nat Rev Neurosci, 6(8):653-659.
Grainger J, HolcombPJ. (2009) Watching the word go by: on the time-course of component processes in visual word recognition. Lang
Linguist Compass, 3(1):128-156.
Grave de Peralta Menendez R, Gonzalez Andino S, Lantz G, Michel CM, Landis T. (2001) Noninvasive localization of electromagnetic
epileptic activity I Method descriptions and simulations. Brain Topogr, 14(2):131–137.
Grave de Peralta Menendez R, Murray MM, Michel CM, Martuzzi R, Gonzalez Andino SL. (2004) Electrical neuroimaging based on
biophysical constraints. Neuroimage,21(2):527-539.
Graves R, Landis T, Goodglass H. (1981) Laterality and sex differences for visual recognition of emotional and non-emotional words.
Neuropsychologia, 19(1):95-102.
Graves WW, Desai R, Humphries C, Seidenberg MS, Binder JR. (2010) Neural Systems for Reading Aloud: A Multiparametric
Approach. Cereb Cortex, 20(8):1799-1815.
Herbert C, Junghöfer M, Kissler J. (2008) Event relatedpotentials to emotional adjectivesduringreading. Psychophysiology, 45:487-498.
Hauk O, Pulvermuller F, Ford M, Marslen-Wilson WD, Davis MH. (2009) Can I have a quick word? Early electrophysiological
manifestations of psycholinguistic processes revealed by event-related regression analysis of the EEG. BiolPsychol, 80(1):64-74.
Iacoboni M, Zaidel E. (1996) Hemispheric independence in word recognition: evidence from unilateral and bilateral presentations. Brain
Lang, 53(1):121-40.
Ilmoniemi RJ, Virtanen J, Ruohonen J, Karhu J, Aronen HJ, Näätänen R, Katila T. (1997) Neuronal responses to magnetic stimulation
reveal cortical reactivity and connectivity. Neuroreport, 8(16): 3537–40.
KisslerJ, Koessler S. (2011) Emotionally positive stimuli facilitate lexical decisions-an ERP study. BiolPsychol,86(3):254-264.
Kissler, J., Herbert, C., Peyk, P., Junghöfer, M. (2007) Buzzwords. Early cortical responses to emotional words during reading.
PsycholSci, 18:475-480.
Koenig T, Stein M, Grieder M, Kottlow M. (2013) A Tutorial on Data-Driven Methods for Statistically Assessing ERP Topographies.
Brain Topogr. Epub 2013 Aug 29.
Koenig T, Kottlow M, Stein M, Melie-GarcíaL. (2011) Ragu: A free tool for the analysis of EEG and MEG event-related scalp field data
using global randomization statistics. ComputIntellNeurosci, 2011:938925.
Kuchinke L, Jacobs AM, Grubich C, Võ ML, Conrad M, Herrmann M. (2005) Incidental effects of emotional valence in single word
processing: an fMRI study. Neuroimage, 28(4):1022-32.
Lalande S, Braun CM, Charlebois N, Whitaker HA. (1992) Effects of right and left hemisphere cerebrovascular lesions on discrimination
of prosodic and semantic aspects of affect in sentences. Brain Lang, 42(2):165-86.
Landis T. (2006) Emotional words: What’s so different from just words? Cortex, 42(1):170-179.
Landis T, Graves R, Goodglass H. (1982) Aphasic reading and writing: Possible evidence for right hemisphere participation. Cortex, 18:
105-112.
Landis T, Regard M, Graves R, Goodglass H. (1983) Semantic paralexia: A release of right hemisphere function from left hemispheric
control? Neuropsychologia, 21:359-364.
138
Lehmann D, Skrandies W. (1980) Reference-free identification of components of checkerboard-evoked multichannel potential fields.
ElectroencephClinNeurophysiol, 48:609-621.
Levy T, Walsh V, Lavidor M. (2010) Dorsal stream modulation of visual word recognition in skilled readers.Vision Res, 50(9): 883-888.
Lidzba K, Schwilling E, Grodd W, Krägeloh-Mann I, Wilke M. (2011) Language comprehension vs. language production: Age effects
on fMRI activation. Brain Lang, 119(1):6-15.
Margulies DS, Vincent JL, Kelly C, Lohmann G, Uddin LQ, Biswal BB, Villringer A, Castellanos FX, Milham MP, Petridesi M. (2009)
Precuneus shares intrinsic functional architecture in humans and monkeys. PNAS, 106(47):20069-20074.
Massimini M, Ferrarelli F, Huber R, Esser SK, Singh H, Tononi G. (2005) Breakdown of cortical effective connectivity during sleep.
Science, 309(5744): 2228–32.
Michel CM, Murray MM, Lantz G, Gonzalez S, Spinelli L, Grave de Peralta R. (2004) EEG source imaging. ClinNeurophysiol,
115(10):2195-2222.
Michel CM, Koenig T, Brandeis D, Gianotti LRR, Wackermann J. (2009) Electrical Neuroimaging. Cambridge (UK). Cambridge
University Press.
Michel CM, Murray MM. (2011) Towards the utilization of EEG as a brain imaging tool. Neuroimage, 61:371-385.
Mohr B, Endrass T, Hauk O, Pulvermüller F. (2007)ERP correlates of the bilateral redundancy gain for words. Neuropsychologia,
45(9):2114-2124.
Mohr C, Michel CM, Lantz G, Ortigue S, Viaud-Delmon I, Landis T. (2005) Brain state-dependent functional hemispheric specialization
in men but not in women. Cereb Cortex, 15:1451-1458.
Morand S, Thut G, Grave de Peralta R, Clarke S, Khateb A, Landis T, Michel CM. (2000) Electrophysiologicval evidence for fast visual
processing through the human koniocellular pathway when stimuli move. Cereb Cortex, 10:817-825.
Murray MM, Brunet D, Michel CM. (2008) Topographic ERP analyses: a step-by-step tutorial review. Brain Topogr, 20:249-264.
New B, Pallier C, Ferrand L, Matos R. (2001) Une base de données lexicales du français contemporain sur internet: LEXIQUE.L'Année
Psychologique,101:447-462.
Ortigue S, Michel CM, Murray MM, Mohr C, Carbonnel S, Landis T. (2004) Electrical neuroimaging reveals early generator modulation
to emotional words. Neuroimage, 21(4):1242-1251.
Papagno C, Fogliata A, Catricalà E, Miniussi C. (2009) The lexical processing of abstract and concrete nouns. Brain Res, 1263:78-86.
Parks NA, Maclin EL, Low KA, Beck DM, Fabiani M, Gratton G. (2012) Examining cortical dynamics and connectivity with
simultaneous single-pulse transcranial magnetic stimulation and fast optical imaging. NeuroImage, 59(3): 2504–10.
PellMD. (2007) Reduced sensitivity to prosodic attitudes in adults with focal right hemisphere brain damage. Brain Lang, 101(1):64-79.
Perrin F, Pernier J, Bertrand O, Giard MH, Echallier JF. (1987) Mapping of scalp potentials by surface spline
interpolation.ElectroencephalogrClinNeurophysiol, 66:75-81.
Ponz A, Montant M, Liegeois-Chauvel C, Silva C, Braun M, Jacobs AM, Ziegler JC. (2013) Emotion processing in words: a test of the
neural re-use hypothesis using surface and intracranial EEG. Soc Cogn Affect Neurosci, In press.
Price CJ. (2012) A review and synthesis of the first 20 years of PET and fMRI studies of heard speech, spoken language and reading.
Neuroimage, 62(2):816-847.
Pylkkanen L, Marantz A.(2003) Tracking the time course of word recognition with MEG. Trends CognSci, 7(5):187-189.
Ragazzoni A, Pirulli C, Veniero D, Feurra M, Cincotta M, Giovannelli F, Chiaramonti R, Lino M, Rossi S, Miniussi C. (2013)
Vegetative versus minimally conscious states: a study using TMS-EEG, sensory and event-related potentials. PloS one, 8(2),
e57069.
Reuterskiöld C. (1991) The effects of emotionality on auditory comprehension in aphasia. Cortex, 27(4):595-604.
Riba J, Heldmann M, Carreiras M, Munte TF. (2010) A neuroimaging study of conflict during word recognition. Neuroreport,
21(11):741-745.
Richardson FM, Seghier ML, Leff AP, Thomas MSC, Price CJ. (2011) Multiple routes from occipital to temporal cortices during
reading. J Neurosci, 31(22):8239-8247.
139
Rochas V, Gelmini L, Krolak-Salmon P, Poulet E, Saoud M, Brunelin J, Bediou B. (2013) Disrupting pre-SMA activity impairs facial
happiness recognition: an event-related TMS study. Cereb Cortex, 23(7):1517-25.
Romei V, Driver J, Schyns PG, Thut G. (2011) Rhythmic TMS over parietal cortex links distinct brain frequencies to global versus local
visual processing. CurrBiol, 21(4):334-337.
Ross ED, Edmondson JA, Seibert GB, Homan RW. (1988) Acoustic analysis of affective prosody during right-sided Wada Test: A
within-subjects verification of the right Hemisphere’s role in language. Brain Lang, 33(1):128-145.
Rossi S, Hallett M, Rossini PM, Pascual-Leone A and The Safety of TMS Consensus Group. (2009) Safety, ethical considerations, and
application guidelines for the use of transcranial magnetic stimulation in clinical practice and research. ClinNeurophysiol,
120(12):2008-2039.
Schacht A, Sommer W. (2009) Time course and task dependence of emotion effects in word processing. Cogn Affect BehavNeurosci,
9(1):28-43.
Scott GG, O'Donnell PJ, Leuthold H, Sereno SC. (2009) Early emotion word processing: evidence from event-related potentials.
BiolPsychol, 80(1):95-104.
Sperry RW. (1961) Cerebral organization and behavior. Science,133:1749-1757.
Springer S, Deutsch G. (2004) Left Brain, Right Brain. Perspectives from cognitive neuroscience. Warsaw (PL).Prószynski I S-ka SA.
Stroobant N, Buijs D, Vingerhoets G. (2009) Variation in brain lateralization during various language tasks: a functional tra nscranial
Doppler study. Behav Brain Res, 199(2):190-196.
Segalowitz SJ, ZhengX. (2009) An ERP study of category priming: evidence of early lexical semantic access. Biol Psychol, 80(1):122129.
Skrandies W. (2013) Electrophysiological correlates of connotative meaning in healthy children. Brain Topogr, Epub 2013 Aug 22.
Spinelli L, Gonzalez Andino S, Lantz G, Seeck M, Michel CM. (2000) Electromagnetic inverse solutions in anatomically constrained
spherical head models. Brain Topogr, 13:115-126.
Stoeckel C, Gough PM, Watkins KE, Devlin JT. (2009) Supramarginalgyrus involvement in visual word recognition. Cortex,45(9):10916.
Thorpe SJ. (2009) The Speed of Categorization in the Human Visual System. Neuron, 62(2):168-170.
Valero-Cabré A, Payne BR, Pascual-Leone A. (2007) Opposite impact on 14C-2-deoxyglucose brain metabolism following patterns of
high and low frequency repetitive transcranial magnetic stimulation in the posterior parietal cortex. Exp Brain Res, 176(4):603-615.
Vigneau M, Beaucousin V, Hervé P-Y, Jobard G, Petit L, Crivello F, Mellet E, Zago L, Mazoyer B, Tzourio-Mazoyer N. (2011) What is
right-hemisphere contribution to phonological, lexico-semantic, and sentence processing? Insights from a meta-analysis.
Neuroimage, 54(1):577-593.
Wagner T, Rushmore, J, Eden U, Valero-Cabré A. (2009)Biophysical foundations underlying TMS: setting the stage for an effective use
of neurostimulation in the cognitive neurosciences. Cortex, 45:1025-1034.
Wernicke C. (1874) Der aphasischeSymptomencomplex: Eine Psychologische Studie auf anatomischer Basis. Breslau (PL). Max Cohn
&Weigert.
Woody CD. (1967) Characterization of an adaptive filter for the analysis of variable latency neuroelectric signals. Med Biol Eng, 5:539553.
Young AW, Bion PJ, Ellis AW. (1980) Studies toward a model of laterality effects for picture and word naming. Brain Lang, 11: 54–65.
Zhang S, Li CR. (2012) Functional connectivity mapping of the human precuneus by resting state fMRI. Neuroimage, 59(4):3548-3562.
140