University of Groningen
The organization of initiation and inhibition of movement
Toxopeus, Carolien Margot
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to
cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2012
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Toxopeus, C. M. (2012). The organization of initiation and inhibition of movement: linking muscle and brain
in healthy subjects and patients with Parkinson's disease Groningen: s.n.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the
author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the
number of authors shown on this cover page is limited to 10 maximum.
Download date: 16-06-2017
CHAPTER
1
General Introduction
9
Chapter 1
Of all living creatures, humans are capable of performing the most astonishing variety of movements
ranging from the simplest, such as retracting the arm in reaction to a painful stimulus, to the finest
and most challenging category exemplified by performing a complex dance choreography. The
latter can even become partly a routine after intense practice. In a normal situation, most of our every
day movements seem indeed to come about automatically. However, even though we hardly realize
this, our movements are the final outcome of a complex organization in which the use of different
muscles has to be adjusted to the requirements of the intended movement. At the cerebral level,
the process of the selection of the appropriate motor output involves various components. In this
process, it is the group of subcortical nuclei referred to as the basal ganglia (BG) that play an eminent
role in movement selection by initiating desired movement while inhibiting undesired movement.
The precise organization of movement initiation and inhibition is not precisely understood. The
main aim of this thesis was to investigate the organization of movement initiation and inhibition
with a focus on the roles of the basal ganglia and interconnected circuitry. By employing several
techniques this thesis studied the link between (i) cerebral organization of movement initiation and
inhibition and (ii) movement parameters at the behavioral level in healthy subjects and patients
with Parkinson’s disease (PD). This patient group was studied in particular because PD-associated
(BG) dysfunction results in disturbed initiation and inhibition of movement.
1.1. Cerebral Organization of Movement
Movement organization is determined, firstly, by the physical characteristics of the musculoskeletal system, including mechanical properties of joints and muscles that constitute the effector
system. The second important determinant of movement organization is the process of movement
planning by the central nervous system (CNS) that enables the initiation of a specific movement
and sends the final motor commands via the peripheral nerves to the effector system. The process
of adjusting different muscles to the requirements of the intended movement involves multiple
levels of movement organization ranging from sensorimotor integration implicated in well-defined
action to the modulation of a new cortically generated motor plan by the basal ganglia (BG). In the
following paragraphs we will further elaborate on these components of movement organization.
1.1.1. Higher order motor control
The cortically generated motor plan is strongly defined by characteristics that match basic
environmental features such as spatial dimension and temporal change. Their integration in a
motor plan is referred to as higher order motor control (Georgopoulos 1991). Given that all of our
movements are situated or related to the external space that surrounds our body, sensorimotor
integration is an essential component of higher-order motor control. Similarly, intrinsic sensory
information concerning the state of the body itself contributes to this motor planning. Intrinsic
sensory information is derived from e.g. proprioceptive feedback and vestibular information, thus
providing the brain with details on the inter-relationship between various body parts (body scheme)
10
Introduction
and the position of the body in relation to the external space. Together, this complex multimodal
sensory information enables the brain to encode the commands for desired movements (de Jong
2011; Kalaska and Crammond 1992; Previc 1998). For example, in order to initiate a purposeful
movement such as reaching for a glass of water, or catching a ball thrown at us, we also have to
take into account multiple spatial characteristics such as location and shape of the target, all within
a mentally created three-dimensional format. This sensorimotor interaction does not concern a
static condition. Following the initiation of intended movement, continuous updating of sensory
information from eyes, ears muscles and skin enables adjustment of enrolled motor programs. E.g.,
while moving the arm and hand towards a moving ball, specific characteristics of the ball such as size
and texture are provided by the sensory system as the ball gets closer. This enables the appropriate
adjustment of the configuration of the hand aperture which facilitates a successful catch (Castiello
2005). In this way, the organization of (purposeful) movement execution may be regarded to arise
from a system that employs both feedback and feedforward mechanisms.
In the last decade, an increasing amount of details became available concerning the involvement
of distinct cortical regions in the cerebral organization underlying sensorimotor integration. Now,
it is well recognized that planning of purposeful movement is largely based on transformation of
sensory information obtained from the environment into external parameters (Kakei, Hoffman,
Strick 2001). Visuomotor transformation is facilitated by parieto-occipital networks (de Jong et al.,
1999; Jeannerod et al., 1995), thereby forming a concept of the desired movement or motor task
which is used as a template for planning of movement by a parieto-premotor network (Binkofski
et al., 1999; Burnod et al., 1999; de Jong, van der Graaf, Paans 2001; de Jong, Leenders, Paans 2002;
Filimon 2010; Wenderoth et al., 2006). Finally, the generation of movement output is primarily
mediated by the primary motor cortex. At this stage, one may expect that more general movement
parameters, such as direction of movement, have to be translated into the selection of specific
muscle activity. The primary motor cortex is known to be partly somatotopically organized, which
has been well documented by e.g. Penfield and Rasmussen’s study of the homunculus, which
forms a schematic representation of different body parts (Penfield and Rasmussen 1950). On the
other hand, the primary motor cortex is not merely a passive somatotopical output station with
a fine-grained representation of individual muscles (Schieber 2001). If that were true, effectively
organizing movements would be very time-consuming and inefficient since timed control of all
individual muscles would provide an enormous number of different combinations in a seemingly
endless number of degrees of freedom (de Rugy 2010). In this respect, the use of information from
the external environment as external parameters, also at the level of the primary motor cortex, may
indeed provide an escape. How the use of external parameters is integrated in the organization of
movement is a question that was already addressed in the late nineteenth century by the renowned
neurologist J.H. Jackson (1835 – 1911) (Jackson 1873). At present, it is hypothesized that in the
primary motor cortex, both intrinsic and extrinsic parameters are used to generate the appropriate
motor output for the intended movement (Post, Bakels, Zijdewind 2009), making the primary motor
11
1
Chapter 1
cortex an important contributor to higher order motor control.
Although the planning and initiation of a movement task is organized at cortical level, before motor
output from the primary motor cortex descends to the brain stem and spinal cord via the pyramidal
tract, several subcortical brain areas, such as the BG and cerebellum, contribute to the modulation
of the cortically generated motor plan resulting in fine-tuned motor output (Groenewegen 2003).
In this process, the selection of the appropriate movement output (initiation) and inhibition of
unwanted movement is of particular importance. In this thesis, the mechanisms of selection are
studied by employing highly specified movement tasks. An associated issue in higher-order motor
control concerns the neuronal mechanisms underlying free selection of one out of multiple possible
motor actions (Beudel and de Jong 2009; de Jong 2011). This topic goes beyond the scope of this
thesis.
1.1.2. Initiation and Inhibition of Movement
Movement initiation and inhibition can be regarded as starting or stopping a specific movement
task, but movement initiation and inhibition can also be regarded at single-joint (effector) level.
At this level, movement initiation and inhibition are defined as the modification of a cortically
generated motor plan by both the selection of appropriate muscles and the inhibition of undesired
muscle activity resulting in coordinated muscle activity. At the level of the effector, movement is not
just ‘on’ or ‘off ’ but instead an orchestration of sequential muscle contractions requiring a balanced
interaction between initiation and inhibition of agonist and antagonist muscles (Raptis et al 2010).
This balance between initiation and inhibition is defined by the purpose of the intended movement
(de Jong and Paans 2007; Hauber 1998). Although the organization of initiation and inhibition is not
quite clear, it is known that the BG, a group of interconnected subcortical nuclei, play an important,
yet, complex role in movement initiation and inhibition. The BG are constituted by the nucleus
caudatus (caudate), putamen and nucleus accumbens (together indicated as the striatum), the
globus pallidus (pallidum), which consists of an external, internal and ventral part, the subthalamic
nucleus (STN), the substantia nigra (SN) and the ventral tegmental area. The BG are connected with
the cortex through the thalamus, in a complex network of parallel loops or circuits supporting
sensorimotor, cognitive and emotional functions (Alexander, Crutcher, DeLong 1990; Parent 1990),
together forming the cortico basal ganglia thalamocortical (CBGTC) circuitry (figure 1.1). The main
role of the BG in this network of different functional domains is thought to be the selection of action
(Redgrave, Prescott, Gurney 1999).
Regarding motor control, the BG receive cortical information from frontal areas through cortical
projections to the striatum, which is the main input station of the BG. The striatum is connected to
the output nuclei of the BG, of which the most important are the substantia nigra pars reticulata
(SNr), and the pallidum interna. The output nuclei mainly project to the frontal cortex via the
thalamus. Besides the connections of the BG with the cortex, there are also connections between
the different nuclei of the BG forming an intrinsic circuitry which is crucial for modulation of the
12
Introduction
1
Figure 1.1 Diagram of basal ganglia pathways
Schematic overview of basal ganglia thalamocortical circuits including the hyperdirect (cortico-subthalamopallidal), direct (cortico-striato-pallidal) and indirect (cortico-striato-GPe-subthalamo-GPi) pathways. Dashed
lines represent the excitatory glutamatergic (glu) projections and solid lines represent the inhibitory GABAergic
(GABA) projections. The dotted arrows indicate nigrostriatal dopaminergic projections; dopamine stimulates
the direct pathway and inhibits the indirect pathway through these projections. STN = subthalamic nucleus,
GPe= pallidum externa; GPi= pallidum interna; SNr, substantia nigra pars reticulate; SNc= substantia nigra pars
compacta (modified from (Groenewegen and van Dongen 2008; Nambu, Tokuno, Takada 2002)).
cortical input that enters the BG (Groenewegen 2003; Mink 1996). The intrinsic circuitry consists
of a direct- and indirect pathway which originate in the striatum and are differentially modulated
by the neurotransmitter dopamine. The direct pathway entails an inhibitory projection between
the striatum and pallidum interna and SNr. The indirect pathway, on the other hand, consists of an
inhibitory projection to the neurons in the external pallidum, and subsequent inhibitory projections
between the pallidum externa and pallidum/SNr, either directly or via the STN (Albin, Young, Penney
1989; Alexander and Crutcher 1990; DeLong and Wichmann 2009; Mink and Thach 1991; Mink and
Thach 1993; Mink 1996). Thus, whereas activation of the direct pathway results in reduced inhibition
of thalamocortical activation and facilitates the cortex in selection of the appropriate motor output,
activation of the indirect pathway increases BG output thereby causing inhibition of unwanted
movement. Regarding movement organization, the net effect of the intrinsic connections between
13
Chapter 1
the BG nuclei is a modulation of cortical input by selecting the appropriate motor output (initiation)
while inhibiting unwanted movement. However, the precise organization of movement initiation
and inhibition by the BG and interconnected circuitry remains unclear. This becomes particularly
evident from patients with pathophysiological changes in the BG: the movement disorders including
Parkinson’s and Huntington’s disease and dystonia. These disorders all have their unique clinical
features which are thought to depend on specific combinations of changes in motor subcircuits of
the BG (DeLong and Wichmann 2007). What the extrapyramidal disorders have in common is that
the balance between initiation and inhibition of movement at effector level is disturbed. In this
respect, PD patients are of particular interest because PD involves a reduction of dopamine which
leads to a changed functionality of the intrinsic BG pathways and causes specific impairments in
movement initiation and inhibition.
1.2. Investigation of Movement Initiation and Inhibition: Relevance to Parkinson’s disease
In neuroscience, when trying to gain insight in the functional organization of a specific area of
the human brain, traditionally, patients with a problem in that brain area are studied. In the ‘preneuroimaging’ era, function and brain area were coupled by linking neurological deficits to a known
lesion. An illustrious example of this ‘deficit-lesion coupling’ is provided by the remarkable case
of the railroad worker Phineas Gage, whose personality changed dramatically after surviving an
accident in which a tamping iron he was using to place explosive powder in holes drilled in rocks
went completely through his skull. Since the accident brought substantial damage to Gage’s (pre-)
frontal cortex leading to major changes in his personality and behavior, this case suggested that the
frontal lobes are involved in emotion, planning and decision-making (Macmillan 2000). The deficitlesion coupling derived from neurological cases provided a global function localization of different
areas of the human brain, such as vision being organized by the occipital cortex, motor control
by the precentral gyrus and sensorimotor integration by the parietal cortex. Nevertheless, in the
last decades in which neurologists and neuroscientists gained access to neuroimaging techniques
that enabled the investigation of the human brain in vivo, it became clear that functions are not
simply topographically localized. Hence, a strict cartography of the brain does not exist and, instead,
different areas of the brain can better be regarded as functional hubs that are part of a network
(Friston 2002; Reijneveld et al., 2007). The study of the cerebral organization of functional modalities
such as motor control should, thus, be focused on different areas of the brain operating together in
a network. In order to understand the functional organization of this network as a whole, however,
one also needs to gain insight in the specific roles of the individual components within this network.
Therefore, the study of patients with pathophysiological changes in one of the components of a
network in order to learn more about the organization of a specific brain function is as relevant
today as it was in the early days of neuroscience. The only difference is that, at present, diseaserelated changes in brain function such as motor control in patients can be studied and compared
with healthy subjects in vivo by using specific movement measures and neuroimaging techniques.
14
Introduction
In PD patients, pathophysiological changes in the functionality of the BG cause specific movement
impairments, involving initiation and execution of movement tasks. By comparing cerebral
activation patterns- and movement parameters related to initiation and inhibition between PD
patients and healthy subjects, more insight in the role of the BG and interconnected circuitry
in movement organization can, thus, be obtained. In a reciprocal way, gaining insight in the
fundamentals of cerebral control underlying movement initiation and inhibition may contribute to
the development of new strategies for therapy in PD (Obeso et al., 2008a; van den Wildenberg et al.,
2006). The next paragraph will provide a brief overview of the pathophysiology of PD in relation to
impaired movement initiation and inhibition.
At the beginning of the 19th century Sir James Parkinson (1755-1824), a surgeon and apothecary from
England, was the first to describe the symptoms of PD as “paralysis agitans”, or, “the shaking palsy” in
his prominent work An Essay on the Shaking Palsy (Parkinson 2002). According to James Parkinson,
paralysis agitans constituted a combination of specific features including rest tremor, decreased
muscle power, an abnormal body posture and a propulsive gait. As a tribute to James Parkinson
for his well-described symptoms of the shaking palsy, we refer to this disease as Parkinson’s disease
which is now regarded as a neurodegenerative disease. Patients usually develop the disease in the
6th decade. The prevalence of the disease in the population over 60 is estimated at 0.5% whereas
the incidence of PD within the total population is 0.15% (Wolters and van Laar 2002). This makes PD,
after Alzheimer’s disease, the most common neurodegenerative disease. The cardinal symptoms
of PD are muscle rigidity, tremor, bradykinesia and abnormalities in posture and balance (DeLong
and Wichmann 2007; Marsden 1994). Additionally, patients may experience autonomic symptoms,
psychiatric symptoms such as depression and hallucinations and cognitive impairments.
The pathophysiological mechanism in PD is a degeneration of pigmented brain stem nuclei, which
includes the dopaminergic substantia nigra pars compacta (SNc), resulting in dopamine depletion
of the striatum, particularly the putamen. Although the cause of this degeneration is not clear it is
suggested that a combination of environmental toxins and genetic predisposition may result in the
destruction of neurons in the SNc and subsequent disposition of α-synuclein containing inclusion
bodies, or, Lewy bodies (Forno 1996). Eventually, the loss of dopaminergic innervations causes
changes in the activity of neural pathways within the CBGTC circuits that control movement (Albin,
Young, Penney 1989; Marsden 1994). When approximately half of the dopaminergic neurons of the
SN pars compacta are degenerated, this results in a shift in the balance between direct and indirect
BG pathways (Wichmann et al., 2011). The net effect of these changes in the balance between
intrinsic pathways is thought to be an excitation of the inhibitory pallidum interna, which in turn
causes an inhibition of the thalamus and results in decreased activation of cortical areas such as
the (pre-)motor cortex (Albin, Young, Penney 1989; DeLong 1990). In this manner the disturbed
balance between indirect and direct pathways in PD patients is thought to account for PD-related
changes in movement execution. These impairments can be task-related, and induce difficulties
15
1
Chapter 1
in the starting and stopping of walking (Giladi 2001), whereas at single-joint level, impaired
movement initiation (Berardelli et al., 1986b) is also associated with insufficient inhibition of the
antagonist muscle (Meunier et al., 2000), a phenomenon which is clinically referred to as rigidity.
The insufficient antagonist inhibition might also cause the impaired performance of purposeful
or gradual movement in PD patients. Although the ‘classical’ model of PD, describing a higher
inhibitory BG outflow to the cortex, provides an explanation for the impaired movement initiation
in PD patients it, however, does not fully explain insufficient movement inhibition and impaired
gradual movement modulation which is also seen in these patients (Obeso et al., 2008b; Redgrave
et al., 2010).To study the normal organization of movement initiation and inhibition and PD-related
changes in this organization, the comparison of movement performance between patients and
healthy subjects provides a first step to gain insight into how changes in cerebrally coded movement
initiation and inhibition influence the execution of movement at effector level in these patients.
Particularly, a dissociation of these basic elements of movement can be obtained by focusing on
relatively simple movements of the wrist.
1.3. Investigation of Movement Initiation and Inhibition: Quantitative Assessment of
Movement Parameters and Cerebral Activation Patterns
In order to investigate the link between cerebral organization and motor output, movement
execution at a behavioral level needs to be quantified by employing specific movement measures,
while cerebral activation patterns can be measured using functional neuroimaging. Traditionally,
one of the most important behavioral measures to study motor control is reaction time (RT), which
represents the time interval between a presented (visual or auditory) stimulus and the subject
reacting to this stimulus, such as by pushing a button. RT during simple movement tasks is thought
to reflect the capability of planning a movement at the cerebral level. In the investigation of motor
control in PD patients RT is often employed as an indication of the ability to initiate a movement task
(Gauntlett-Gilbert and Brown 1998). For the study of impaired movement execution in PD patients,
movement execution also needs to be described quantitatively. Movement execution can be
quantified by kinematic variables which are thought to reflect the ability to adjust muscle activity to
requirements of the movement task (Almeida et al., 1995; Brown and Cooke 1981; Pfann et al., 2004).
In addition, kinematic variables are linked to cerebral activations of the motor circuitry (Turner et
al., 2003). Furthermore, movement execution can be quantitatively described by muscle activity
employing electromyography (EMG) and specific EMG parameters can be used to further investigate
the link between impaired modulation of initiation and inhibition at the levels of cerebrally coded
task performance and execution by the musculoskeletal effector system.
The following paragraphs will be used to elaborate on the various techniques that were employed
in this thesis to study the organization of movement initiation and inhibition at behavioral,
neurophysiological and cerebral level in PD patients and healthy subjects.
1.3.1. Quantification of Movement Execution: Kinematics
16
Introduction
The description of motion of objects using the displacement of an object as function of time,
without reference to the forces causing the displacement, is referred to as kinematic description
of motion (An and Chao 1984; Enoka 2002). The study of kinematics may concern any rigid object
in motion but in the context of human movement science, or biomechanics, the ‘object’ in motion
may be a human body or an extremity such as an arm or hand performing a specific movement
task. Typically, kinematic measures are used to determine motion features within a relatively simple
coordinate system such as the Cartesian coordinate system. The most important kinematic variables
are (joint) position and displacement -which can be linear or angular and are measured relative to
a baseline value or location within space-, movement velocity and movement acceleration. Initially,
kinematic variables were used in combination with X-rays to gain insight in the mechanic aspects of
the wrist bones and normal joint function (von Bonin 1929). Nowadays, kinematic analysis is applied
to quantify, for example, functional aspects of goal-directed or purposeful wrist/hand movement.
Moreover, since kinematics are thought to reflect movement organization at a central level, these
variables are widely used to investigate human motor control (Ramos et al., 1997). An important
finding was that even though theoretically, there is an endless number of ways to make a particular
movement, movement trajectories seem to be performed with specific, invariant kinematic
characteristics (Abend, Bizzi, Morasso 1982). A good example of these kinematic invariants is
provided by the bell-shaped velocity curves of goal-directed movements (Georgopoulos, Kalaska,
Massey 1981; Plamondon et al., 1993; Soechting and Lacquaniti 1981). It is thought that these
velocity curves are bell-shaped due to delays in responding to command parameters of the various
components of the motor system such as sensory-feedback and (synergistic) activity of different
muscles that are coupled. Additionally, another important kinematic invariant is the speed-accuracy
trade-off ; a phenomenon which describes that movement accuracy can be achieved at the cost of
speed (Elliott, Helsen, Chua 2001; Fitts 1992). Given that kinematic profiles of diverse (goal-directed)
movements have invariant characteristics and, therefore, seem to answer to specific physical laws,
many studies aimed to compose a model describing these profiles (Meyer et al., 1988; Plamondon et
al., 1993). These models try to take into account the many linearly and serially coupled components
the motor system comprises such as muscles and nerves. Furthermore, from the shape of kinematic
profiles it can be derived how well neuromuscular components of the motor system are tuned and
work together as synergists, thereby obtaining optimal movement performance. To quantify optimal
movement control a measure that is often used is movement variability, which is hypothesized to
be the result of a movement-to-movement variance of neuromuscular components (Plamondon
and Alimi 1997). Thus, a low variability of movement is thought to reflect optimal movement
control. Since kinematic variables such as movement variability reflect the neural system’s ability to
optimize motor performance, these variables are also used to investigate detoriation of movement
organization as a result of age-related changes (Yan 2000) and impaired movement control in
movement disorders such as PD (Sheridan, Flowers, Hurrell 1987).
1.3.2. Quantification of Muscle Activity: Electromyography
17
1
Chapter 1
Electromyography (EMG) which means “writing of electrical aspects of muscles” is a method to study
muscle functionality employing the electrical signals generated by action potentials travelling
along the muscle fibers during contraction (Basmajian and De Luca 1985; van Putten 2009). The first
observation of a relationship between muscle contraction and electrical current was made by the
Italian physician and physicist Luigi Galvani (1737-1789), in muscles of various animals (figure 1.2).
Galvani published the results of his experiments in De Viribus Electricitatis in Motu Musculari
Commentarius and his work is regarded as the foundation of neurophysiology (Piccolino 1998).
The knowledge of the relationship between electrical current and muscle contraction was further
Figure 1.2: Schematic drawing of Galvani’s frog leg experiment (www.wikipedia.org)
developed to measure signals of human muscles in the first half of the 20th century. At present, EMG is
clinically commonly used to detect neuromuscular changes whereas for research applications, EMG
is used to investigate patterns of skeletal muscle activity in the context of movement organization in
both healthy subjects and patients with neuromuscular diseases or movement disorders.
The coordinated activation of skeletal muscles facilitates our body to move in a functional fashion.
Skeletal muscles consist of a large amount of muscle fibers, which are independently functioning
as contractile components. Muscle fibers are controlled by a motor neuron, which is a cell body
located in the ventral horn of the spinal cord or brainstem that is connected to the muscle fibers by
a descending motor axon (Basmajian and De Luca 1985; Kandel and Schwartz 2000). Together, the
18
Introduction
motor neuron and the muscle fibers it innervates are referred to as the motor unit and it depends
on the type and location of the muscle how many fibers are controlled by the same motor neuron.
The motor axon controls the muscle fibers via the so-called motor end-plate which facilitates that
signals from the central- and peripheral nervous system lead to activation of the skeletal muscles.
The motor end-plate is a synaptic structure that enables communication between the motor neuron
and the muscle fiber by chemical signaling. This way an all-or-none action potential (facilitated by
the exchange of Na+ and K+ ions across cell membranes) can be generated along the sarcolemma
causing the fibers to contract. In case of a coherent contraction of the majority of muscle fibers,
this induces movement. EMG measures the combined action potentials of motor units that cause
a change in the polarity (positive or negative) of membranes. This produces a myoelectric signal
with an amplitude in the order of 100 milli Volt (V) and frequencies ranging from 0 to 500 Hz.
When measured during a specific time window and/or during a movement task, the EMG signal
of a muscle reflects the dynamic activity pattern of that particular muscle. Since muscle activity is
controlled by selective activation and inhibition by the nervous system through the motor neurons,
EMG activity patterns reflect how appropriate muscles are selected and provide further insight in
movement organization (Hermens et al., 1999; Reaz, Hussain, Mohd-Yasin 2006).
There are two different types of EMG recording; needle EMG to measure muscle activity
intramuscularly and surface EMG to measure muscle activity by attaching electrodes with gel or
adhesive strips to the skin overlying the target muscles (Pullman et al., 2000). Needle EMG is more
precise, since it measures the activity of a few adjacent muscle fibers, but has a limited field of view.
Surface EMG, on the other hand measures the global electrical potential of multiple motor units
(Farina, Merletti, Enoka 2004). However, there is a chance that some of the signal detected over a
muscle might be generated by a nearby muscle (Basmajian and De Luca 1985; De Luca and Merletti
1988; Winter, Fuglevand, Archer 1994). The latter phenomenon is referred to as crosstalk and causes
surface EMG to be regarded a less accurate method to investigate the activity of specific muscles,
although the spatial resolution of surface EMG is enhanced by employing multichannel EMG (Drost
et al., 2006; Zwarts and Stegeman 2003). Additionally, another factor that contributes to surface
EMG being less accurate than needle EMG is the resistance of the skin, which causes a lower signal
to noise (S/N) ratio although the S/N ratio can be increased by abrasion and shaving of the skin prior
to placing the EMG electrodes and the use of saline paste or gel to attach the electrodes (Hermens
et al., 1999). On the other hand, surface EMG, as opposed to intramuscular EMG, has the advantage
of being non-invasive and allowing longer recordings, which makes it a preferred tool to use for
research applications.
In order to measure surface EMG (further referred to as ‘EMG’), electrodes are used that typically
consist of silver/silver chloride (Ag/AgCl) and have shapes varying from circular to rectangular.
The electrodes are placed over the belly of the muscle, which is defined as the part between the
insertion and origin of the muscle with the thickest diameter. Surface EMG is usually recorded with
electrodes in a bipolar configuration with a recommended distance between the electrodes of 1-2
19
1
Chapter 1
cm (Basmajian and De Luca 1985; Hermens et al., 1999). In this configuration, two potentials are
detected, each with respect to a reference electrode. An important reason to employ the bipolar
configuration is that by subtraction of the signals measured by these two electrodes a large part of
the noise component produced by nearby muscles is removed, thereby reducing cross-talk.
After several preprocessing steps, the EMG can be analyzed as a function of time (time-domain
analysis) to identify patterns of co-active muscles during movement (Soderberg and Knutson 2000)
or the frequencies of the EMG signal can be analyzed (frequency-domain analysis). Moreover, when
EMG signals are analyzed in combination with kinematic parameters, the relation between muscle
activity patterns and movement performance can be investigated.
1.3.3. Measuring Cerebral Activation patterns: functional Magnetic Resonance Imaging
Magnetic resonance imaging (MRI) is a widely used non-invasive technique to study human anatomy
in vivo. MRI employs the differences in magnetic properties between various types of tissues of the
human body to create an anatomical image (Smith and McCarthy 1992). Measuring the behavior of
hydrogen atoms is accomplished by first aligning the atoms in the static MR field, thereby obtaining
a state of the hydrogen atoms called longitudinal magnetization or equilibrium. Next, the hydrogen
atoms are excited by applying a radio frequency (RF) pulse. When turning this pulse off, the hydrogen
atoms return to their previous state, the equilibrium. This return is accompanied by the emission of
energy (resonance) and the differences in the timing of the return to equilibrium of hydrogens in
different tissues cause the differences in signal intensity which constitute the MR image. In addition
to structural images, MRI can also generate functional images of the brain; functional Magnetic
Resonance Imaging (fMRI).
fMRI was first used in the early nineties (Kwong et al., 1992; Ogawa et al., 1992). Although fMRI can
not be used to measure the activity of single neurons, it does provide an index for regional neuronal
activity. Therefore, fMRI provides a rich tool to investigate human brain function non-invasively
(Raichle 1998). Instead of directly measuring neural activity, such as in electroencephalography
(EEG), fMRI determines brain activation by using the relation between blood flow and neuronal
activation. This relation is based on neuronal activation being accompanied by an increment of
blood flow, a phenomenon already found in 1890 by Roy and Sherrington (Roy and Sherrington
1890). This increment is described by the hemodynamic response function (HRF). The increase
of blood flow is associated with an increase in the supply of particularly oxygenated blood
(oxyhemoglobine), which is relatively larger than the regional extraction of oxygen in the activated
neuronal tissue (Fox and Raichle 1986). Oxygen extraction causes oxyhemoglobine to turn into
deoxyhemoglobin. Since oxyhemoglobine is diamagnetic and increases the MR signal whereas
deoxyhemoglobin is paramagnetic causing a weaker MR signal, changes in the amounts of oxy- and
deoxyhemoglobine modify the magnetic field as experienced by the protons in water molecules
surrounding these protons. These changes in the magnetic field affecting the protons constitute
the measured magnetic resonance signal (Ogawa et al., 1990). The signal that is obtained from the
20
Introduction
regional differences in oxygenated blood is also called the Blood Oxygenation Level Dependent
(BOLD) signal. By dividing the brain in three-dimensional pixels (voxels), local changes in the BOLD
signal can be compared statistically by deriving t-values per voxel. When employing fMRI during
specific (motor) tasks, these t-values can be used to investigate differences in activations between
tasks and/or groups of subjects using, for example, statistical parametric mapping (SPM) (Friston
et al., 1995) or other fMRI analysis software packages. These analyses produce statistical images, or
T-maps. To identify the anatomical locations of foci of activation, these T-maps can be rendered on
probabilistic maps using software such as MRICron (Rorden, Karnath, Bonilha 2007).
Although the fMRI activation maps are interpreted as a reflection of regional neural activity related
to, for example, the execution of a specific motor task, the exact relationship between neural activity
and increased blood flow is not completely understood and it is debated whether increase of blood
flow is related to a regional use of energy or related to neural signalling mechanisms (Attwell and
Iadecola 2002; Kleinschmidt and Muller 2010; Logothetis et al., 2001). Furthermore, the temporal
resolution of fMRI (a few seconds), is lower compared to other investigational methods such as EEG.
On the other hand, fMRI has a relatively high spatial resolution (in the order of a few millimeters)
and in contrast with EEG, fMRI can be used to directly investigate task-related activation patterns in
structures that are located deep in the brain. The latter makes fMRI a useful method to investigate
cerebral activation patterns related to movement initiation and inhibition in smaller, subcortical
brain areas such as the BG.
1.4. Outline of this Thesis
The main topic of this thesis concerns the organization of initiation and inhibition of movement
with a focus on the basal ganglia and interconnected circuitry. This organization was investigated in
healthy subjects and patients with Parkinson’s disease (PD) by employing several hand movement
tasks and a range of different techniques measuring motor control at the cerebral level (fMRI) as well
as quantitatively assessing movement parameters using kinematic, and neurophysiological (EMG)
measures.
In chapter 2 we identified the normal activation patterns related to movement initiation and inhibition
by employing fMRI and three different hand movements dissociating between movement initiation,
inhibition and smooth modulation of movement in healthy subjects. In addition to the investigation
of movement organization at a more ‘simple’ level in chapter 2, we aimed to further elucidate how
movement at a higher cortical level is organized in healthy subjects in chapter 3. To that end, we
used a multidirectional manual step-tracking task to explore how higher-order information defining
the direction of hand movement, is incorporated in the somatotopic representation of the manual
effector system in the human primary motor cortex.
Whereas chapters 2 and 3 focus on normal organization of initiation and inhibition of movement
and higher order motor control at cerebral level, the next chapters are focused on the changes in
the organization of movement initiation and inhibition in PD patients. As a first step, we investigated
21
1
Chapter 1
whether the use of a wrist manipulandum in combination with a specific movement task is a
valid method for further study of motor control in PD patients in chapter 4. While for initiation
and inhibition of movement the effect of simultaneous contraction and relaxation of antagonist
muscle groups is quite apparent, smooth execution of purposeful movement, which is one of the
main goals of motor control, is achieved more subtly by gradual modulation of muscle synergists.
In PD patients not only abrupt initiation and inhibition of movement is affected: these patients also
experience problems executing purposeful movement. Therefore, we also studied the organization
of gradually modulated (smooth) movement, requiring a balanced control of muscle synergies. In
chapter 5 we explored whether a task consisting of continuous wrist circumduction, using the wrist
manipulandum, can be used to quantify smooth movement execution. To that end, we investigated
the relation between muscle activity and gradual movement performance by using both kinematic
and electromyographic (EMG) recordings and comparing results between healthy elderly and
young adults. Since the continuous wrist circumduction task was found to be sensitive to changes
in control of gradual movement at the behavioral level, in chapter 6 we also used this task to
investigate differences in continuous movement requiring gradual muscle adjustment between
Parkinson patients and healthy subjects.By employing fMRI and the different movement tasks
explored in the previous chapters, chapter 7 explores PD-related changes in cerebral activation
patterns associated with initiation, inhibition and gradual movement modulation at a cerebral
level. The used tasks involved abrupt movement initiation and inhibition in simple movements
as well as more complex movements requiring synergistic muscle activities. Furthermore, specific
comparisons between these movement tasks were made to enable dissociation between BG
activations related to modulation of synergistic muscle activity underlying movement execution
and the enhanced demand of visuomotor transformations.
The results in chapter 5-6 indicated that smooth movement requires gradual transitions of opposed
muscle activity, while the findings of chapter 7 indicated that the BG are also involved in smooth
modulation of movement. Regarding the role of the BG in the organization of movement that
demands tuned coordination of co-activate muscles, in chapter 8 we hypothesized that in the
BG of PD patients impaired selection of appropriate movement output may become particularly
evident. This was accomplished by using the step-tracking task from chapter 3 and dividing the
different movement directions into ‘singular’ movement directions, dominated by the activity of
one agonist muscle, and ‘composite’ movement directions, involving a higher degree of muscle coactivity tuning. Activations related to these two types of directions were compared between healthy
subjects and patients at the behavioral, neurophysiological and cerebral level.
Finally, in chapter 9 the results of the chapters 2-8 are discussed in a broader perspective and
concludes these chapters 2-8. Chapter 10 provides future perspectives in the investigation of the
organization of movement.
22