Visual Pathways for Object-Oriented
Action and Object Recognition: Functional
Anatomy with PET
Isabelle Faillenot, Ivan Toni, Jean Decety, Marie-Claude
Grégoire1 and Marc Jeannerod
The purpose of this study was to identify the functional anatomy of
the mechanisms involved in visually guided prehension and in object
recognition in humans. The cerebral blood flow of seven subjects
was investigated by positron emission tomography. Three conditions
were performed using the same set of stimuli. In the ‘grasping’
condition, subjects were instructed to accurately grasp the objects.
In the ‘matching’ condition, subjects were requested to compare the
shape of the presented object with that of the previous one. In the
‘pointing’ condition (control), subjects pointed towards the objects.
The comparison between grasping and pointing showed a regional
cerebral blood flow (rCBF) increase in the anterior part of the inferior
parietal cortex and part of the posterior parietal cortex. The
comparison between grasping and matching showed an rCBF
increase in the cerebellum, the left frontal cortex around the central
sulcus, the mesial frontal cortex and the left inferior parietal cortex.
Finally, the comparison between matching and pointing showed an
rCBF increase in the right temporal cortex and the right posterior
parietal cortex. Thus object-oriented action and object recognition
activate a common posterior parietal area, suggesting that some
kind of within-object spatial analysis was processed by this area
whatever the goal of the task.
experimental data have suggested that parietal cortex might also
be involved in some aspects of shape discrimination. Electrophysiological studies in the monkey have shown the existence, in
the anterior intraparietal area (AIP), of neurons which fire in
anticipation of manipulative hand movements directed at
specific objects (Taira et al., 1990; Sakata and Taira, 1994).
Transient inactivation of this area abolishes the preshaping of
the hand during prehension (Gallese et al., 1994).
Neuropsychological observations also point to a similar
organization of visual cortical pathways in man. Parietal lesions
produce spatial disorientation, especially when located in the
right hemisphere. A subset of these lesions centered in the
intraparietal sulcus (IPS) produce a specific visuomotor disorder
known as ‘optic ataxia’. Patients with optic ataxia are unable to
appropriately orient their hand when reaching out towards
target objects (Perenin and Vighetto, 1988), or to correctly
calibrate their finger grip to the size of these objects (Jeannerod,
1986). Conversely, whereas parietal patients have no difficulty in
discriminating between the same objects, patients with bilateral
occipito-temporal lesions present object agnosia (see review in
Farah, 1992). Goodale et al. (1991) reported the case of one such
patient who was unable to visually identify objects but had
normal performance in grasping them. These observations have
prompted a re-evaluation of the functional distinction between
the two cortical visual streams (e.g. Goodale and Milner, 1992).
Both streams would process visual information about object
orientation and shape, but each stream would use it in order to
achieve different goals. The ventral stream plays a critical role in
the visual perception of objects while the dorsal stream mediates
the sensorimotor transformations required for visually guided
action directed towards them.
The present experiment was designed to identify the neural
network involved in object analysis when it is performed with
the purpose of either recognizing the object or generating an
action towards it. Subjects were thus instructed either to make a
shape judgement on non-verbalizable objects of different shapes
and sizes, or to grasp and manipulate them. These two tasks are
not equivalent, however, because making a shape judgement — a
perceptual task — is devoid of any motor component. For this
reason, in order to compare the two tasks, we used pointing as a
third condition for substracting from grasping as much of its
motor component as possible.
A large body of anatomical data indicate that connections within
the monkey visual cortex are organized along two main streams
of projections. Both streams originate from the primary visual
areas: the dorsal stream goes through extrastriate visual areas
and reaches the associative parietal areas, whereas the ventral
stream reaches the inferotemporal cortex (Mishkin et al., 1983;
Boussaoud et al., 1990). Although these pathways can be seen as
two separate routes for processing visual input, they are also
interconnected (Morel and Bullier, 1990).
The point of whether these two visual pathways correspond
to distinct and complementary visual functions has been the
subject of many physiological studies. The inferotemporal cortex
has been known for a long time to be involved in shape
discrimination (Gross et al., 1972; Pohl, 1973). Neuronal
populations in these areas respond more to complex visual
patterns than to simple bars or spots (Tanaka et al., 1990;
Tanaka, 1993). Many cells exhibit remarkable specificity for
particular stimuli, like faces, without being sensitive to more
elementary attributes of these stimuli, such as spatial position,
size, luminance, color or spatial frequency (Perrett et al., 1982,
1987). This property is a strong argument for the claim that
inferotemporal areas are functionally involved in processing
invariant aspects of visual stimuli and thus represent the final
stage of an object identification mechanism.
Posterior parietal cortex, by contrast, has been assigned a role
in processing spatial relationships between objects and between
the body and other objects. Lesions of this region lead to
disturbances in discriminating the respective positions of
objects and in orienting movement in extrapersonal space (Pohl,
1973; Faugier-Grimaud et al., 1978). More recently, however,
Vision et Motricité, INSERM U 94, 16 Av du Doyen Lépine,
69500 Bron and 1CERMEP, 59 bld Pinel, 69003 Lyon, France
Materials and Methods
Subjects
Seven healthy male volunteers (age 24 ± 1 years) were studied. All
subjects were right-handed according to the Edinburgh Handedness
Inventory (Oldfield, 1971). None of them had any history of neurological
disease. Subjects gave their informed consent for the experimental
procedures and were paid for their participation. The experiments were
Cerebral Cortex Jan/Feb 1997;7:77–85; 1047-3211/97/$4.00
carried out in accordance with the Declaration of Helsinski and with
approval from the local Ethical Committee.
Experimental Design and Apparatus
The subjects were studied during three different activation tasks repeated
twice in a pseudo-randomized order. The following specific instructions
concerning each task were given immediately prior to each scan:
.
.
.
Point with the right index finger towards the center of the object and
come back to the starting position (pointing task).
Take the object with the right hand and place it on the table (grasping
task). Subjects were asked to grasp the objects between their
fingertips (precision grip) and to place at least one finger in the notch
(see Fig. 1) in order to adapt their hand shape and orientation to the
object.
Observe the objects and press a mouse button each time the shapes of
two consecutive objects are identical, irrespective of size and
orientation (matching task). Each subject’s right hand was placed in
contact with the mouse for the duration of the task.
The same set of stimuli was presented at the same rate (0.3 Hz) and in
the same pseudo-randomized order in the three conditions. The stimuli
consisted of white wooden blocks (1.0–2.5 mm thick) with five different
shapes and three different sizes (Fig. 1). Each object was presented three
times (45 presentations). The sequence was arranged in such a way as to
involve 15 occurrences of consecutive identical shapes. Objects were
designed in order to elicit different patterns of grasp movements and were
sufficiently similar to require careful visual analysis to discriminate
between each other.
The object presentation device was positioned on the camera-bed. It
consisted of two planes, an inferior horizontal table and a superior
oblique board with a circular hole (13 cm diameter). The subject’s right
forearm was placed on the horizontal table, above the subject’s chest. The
superior board was positioned in such a way that the hole was at reaching
distance and perpendicular to the subject’s gaze axis. Objects were fixed
on a disk (14 cm diameter) by Velcro and presented one at a time through
the hole. In this way, subjects could only see the stimulus upon a black
background. The object notch was always presented on the right side, but
as objects were manually fixed on the disk, small variations in the
orientation of the object could occur. This is why the subjects were
instructed not to pay attention to the orientation. The movements
performed by the subjects towards the objects consisted of rotation of the
arm at the shoulder joint.
Data Acquisition
Regional cerebral blood f low (rCBF) was measured by recording the
distribution of cerebral radioactivity following injections of H215O. Any
increase in rCBF entails an increase in the amount of radioactivity
observed in that region (Raichle, 1987). Each subject received six i.v.
bolus injections of 50 mCi with a 15 min inter-scan interval. Head
movements were restricted with a molded-foam head holder, which was
individually made for each subject and used for both magnetic resonance
imaging (MRI) and positron emission tomography (PET) scans.
Anatomical images (FLASH 3D, T1) were obtained with a Siemens 1.5
Tesla scanner. Using MRI references, PET data were acquired parallel to
the intercommissural (AC–PC) plane. PET scans were performed on a
TTV03 time-of-f light camera (LETI). This scanner provides images with
seven slices (Mazoyer et al., 1990). Since the axial field of view of this
camera (84 mm) is smaller than the whole brain size (∼110 mm), two
acquisitions were required for each condition. This was performed by
moving the bed of the scanner in order to cover the whole cerebrum.
Radiation attenuation phenomenon was corrected by transmission scans
using an external positron emitting source (68Ge) rotating around the
subject’s head.
Bolus injection, data acquisition and object presentation started at the
same time. Data were acquired in list mode and a dynamic series was
generated to plot activity as a function of time. Off-line reconstruction
started when 20% of maximal activity was reached, on average 18 s after
the injection. Emission data were reconstructed for 90 s using a Hanning
filter with a cut-off frequency of 0.15 mm–1. Initial images were in a 128 ×
78 Cerebral Regions Involvement in Grasping and Shape Matching • Faillenot et al.
Figure 1. Object shapes used in the tasks. Each shape was built in three sizes.
128 × 7 pixel format with a 2 × 2 × 12 mm voxel size. MRI data were
converted in a 128 × 128 format with cubic voxel of 2 mm3.
Image Transformations
Image calculations and manipulations were carried out on Sun SPARK
workstation using the ANALYZE image display software and SPM package
running on MATLAB.
PET scans were normalized for global f low for each subject. Possible
head movement between scans was corrected by aligning all scans on the
first one, using Automated Image Registration (AIR) software (Woods et
al., 1993). For each subject, the two sets of PET images were aligned on
MRI data in order to be parallel to the AC–PC plane. After merging, a
unique set of PET images representing the whole brain was available for
each condition.
PET and MR images were transformed into a standard space (Talairach
and Tournoux, 1988) using a stereotactic normalization developed by
Friston et al. (1989). This normalizing spatial transformation matches
each scan to a template image that already conforms to the standard
space. The procedure involves an affine (linear) and quadratic (nonlinear)
three-dimensional transformation. This is followed by a two-dimensional
piecewise (transverse slices) nonlinear matching. MR images were
transformed in the same manner and were averaged across subjects. The
blurring in the mean MRI scan ref lects the variability in position of
anatomic structures for our group.
Subsequently, the PET images were filtered with a low-pass Gaussian
filter (FWHM of 12 mm) for smoothing the data in three dimensions. This
allowed the suppression of noise and effects due to residual differences in
functional and gyral anatomy during intersubject averaging.
Statistical Parametric Mapping
The task and subject effects were estimated according to the general
linear model at each and ever y voxel. To test the hypothesis about
regional specific task effects the estimates were compared using linear
contrasts. The resulting set of voxel values for each contrast provided a
statistical parametric map [SPM (Z)]. In this study, the resulting foci were
then characterized in terms of peak height (u). The significance of each
region was estimated in terms of the probability that the peak height
observed could have occurred by chance [P (Zmax > u)] over the entire
volume.
Stimulus-induced changes in rCBF were considered to be significant at
a threshold of P < 0.05 (Z-score > 4.2) after correcting for multiple
comparisons. Significant rCBF increases were superimposed on averaged
MRI. This procedure enabled us to report activated foci in terms of
Talairach and Tournoux coordinates as well as by reference to anatomical
structures.
Singular Value Decomposition (SVD)
In addition to the inferential statistical analysis described above, a
descriptive statistical analysis concerning the pattern of correlated
activity of PET data was also performed. Principal components of the
covariance matrix were extracted and maps of functional networks
corresponding to significant components were analysed (Friston, 1994).
SVD reveals the spatial and/or temporal patterns that explain most of the
variance of the data.
Results
Behavioral Observations
In the pointing task, the subject’s hand was placed with the
palmar surface resting on the table. During the pointing movement, subjects extended their index finger and f lexed the other
fingers.
In the grasping task, subjects effectively adapted their hand
posture to the different objects by changing wrist orientation,
and adjusting grip size and the number of fingers involved.
In the matching task, although all subjects reported some
difficulty in sustaining attention, the mean error rate was only
6%. Subjects’ strategy for matching objects was based primarily
on global shape and secondarily on shape details. Two subjects
mentioned that they had tried to name the objects. None of them
reported having used mental rotation to match the shapes.
Statistical Parametric Mapping
The Talairach and Tournoux atlas coordinates of all significant
rCBF increases are shown in Table 1, together with the corresponding anatomical regions and Brodmann areas (BA).
Matching Task
When the matching task was compared with the pointing task,
three foci of rCBF increases appeared in the right hemisphere
(Fig. 2, upper part): one in the temporal lobe and two in the
parietal lobe. In the right temporal lobe a discrete but highly
significant rCBF increase was located just above the
temporo-occipital sulcus in the inferior temporal cortex (BA 37).
In the right parietal lobe the rCBF increases were both located in
the dorsal cortex around the intraparietal sulcus (IPS). The lower
region was in the posterior part of the superior parietal lobule
(between BA 19 and 7). The upper region was in the IPS
(between BA 40 and 7).
When the matching task was compared with the grasping task
(Fig. 2, lower part), two rCBF increases appeared. The right
temporal activation observed in the previous subtraction was
still present. The other rCBF increase was in the left middle
frontal cortex (BA 10). The dorsal parietal activations observed
in the previous subtraction were not significant.
Grasping Task
Comparing the grasping task with the pointing task revealed
only one region of activation in the whole brain. The Talairach
and Tournoux atlas coordinates of this rCBF increase
corresponds to BA 40 in the left hemisphere (contralateral to the
grasping hand) extending from AC–PC +16 to +32 mm (Fig. 3).
Careful inspection of the averaged and individual MR images
revealed that this area was located in the inferior part of the
post-central sulcus (BA 2/40).
The comparison between grasping and matching shows the
ensemble of regions involved in the grasping movement, as
illustrated in Figure 4. Planes below AC–PC –8 mm demonstrate
activation in the vermis and in the right lateral cerebellum. The
planes +40 to +48 mm show activation in the left frontal cortex
around the central sulcus (BA 1–4), which extends to the
pre-central sulcus (BA 6). This region covers the sensorimotor
hand area (AC–PC +40 to +48 mm) and the lateral premotor
cortex. The planes +48 and +52 mm show an rCBF increase in the
mesial frontal cortex (BA 6), corresponding to SMA.
Moreover, there are activations in the left inferior post-central
sulcus (BA 2/40) (AC–PC +20 to +32 mm), in the left SII (AC–PC
+8 to +16 mm), in the left insula (AC–PC +0 to +8 mm) and in the
Table 1
Anatomical regions that showed significant rCBF increase
Subtraction Anatomical region
Brodmann area Coordinates
x
Matching vs R intraparietal sulcus
pointing
R occipito-parietal cortex
R inf. temporal gyrus
Matching vs L mid. frontal gyrus
grasping
R inf. temporal gyrus
Grasping vs L inf. postcentral sulcus *
pointing
Grasping vs vermis
matching
R cerebellum
L inf. postcentral sulcus*
L parietal operculum* (SII)
L post. insula *
L post-central gyrus
L central gyrus
L mesial frontal cortex
L cuneus
R post-central sulcus
y
Z-score
z
40/7
34
–52
44
4.80
19/7
37
24
44
–72
–46
40
–8
4.78
4.25
10
–26
60
8
4.81
37
44
–46
–8
4.60
–48
–18
16
4.64
–50
–28
28
4.43
0
–66
–16
5.37
12
–48
–42
–36
–32
–30
0
–16
38
–56
–22
–22
–28
–32
–22
–2
–76
–26
–16
24
8
4
48
48
48
16
32
4.58
5.13
4.47
4.41
4.71
4.57
4.55
4.49
4.42
2/40
2/40
1. 2. 3
4
6
18
2/40
Coordinates are in millimeters and correspond to the atlas of Talairach and Tournoux (1988).
Brodmann areas are defined according to the same atlas. Z-score > 4.2 and > 4.6 correspond to
P < 0.05 and < 0.01 respectively. Post, inf, mid, R and L indicate posterior, inferior, middle, right
and left respectively.
*Peaks of activity localized according to their position on averaged MRI (see figures). The
coordinates do not correspond to these anatomical regions in the atlas.
left cuneus (BA 18) (AC–PC +16 to +20 mm). In addition, an
activated area was found in the fundus of the ipsilateral (right)
post-central sulcus (BA 2/40) (AC–PC +32 mm).
The limited zones in the occipito-parietal cortex and in the IPS
that appeared to be activated when the matching task and the
pointing task were compared (Fig. 2) did not reach the 0.05 level
of significance when comparing the matching task to the
grasping task and the grasping task to the pointing task. In fact,
the maximal Z-scores of these areas was just below the
significance level. IPS showed an rCBF increase in the grasping
versus pointing comparison (Z-score = 4.13, P = 0.06,
coordinates: 30, –52, 44) whereas the occipito-parietal cortex
showed an rCBF increase in the matching versus grasping
comparison (Z = 4.09, P = 0.07, coordinates: 24, –72, 40).
Because the SPM analyses are inherently conservative, these
results suggest that the IPS is likely to be involved in both
matching and grasping, whereas the occipito-parietal cortex is
more distinctive of object identification.
Functional Networks
The SVD analysis showed that the main factor accounts for 77.5%
of the observed variance–covariance structure (Fig. 5A). Figure
5B presents the score for this factor, revealing that the main
effect is present in the grasping and matching conditions. In
other words, the PET activity observed in the present
experiment was distributed between two poles corresponding
to our two main experimental conditions. Therefore, the
pointing task can be considered as a control condition for the
other two tasks.
The eigenimages corresponding to the main factor are shown
in Figure 5C and D. The positive eigenimage (Fig. 5C) represents
Cerebral Cortex Jan/Feb 1997, V 7 N 1 79
Figure 2. Cerebral regions activated during the matching task compared with the pointing task, and during the matching task compared with the grasping task. These regions are
superimposed on sections of the averaged MRI. Under horizontal slices, numbers indicate the distance (in mm) from the AC–PC plane. The red and yellow areas indicate pixels with
Z-score > 3.1 and > 4.2 (P corrected < 0.05) respectively. The right hemisphere is on the right. VAC is the vertical traversing the anterior commissure.
80 Cerebral Regions Involvement in Grasping and Shape Matching • Faillenot et al.
Figure 3. Cerebral regions activated during the grasping task compared with the pointing task. All conventions are as for Figure 2.
the functional network in the grasping task. This network
includes principally the cerebellum and the left sensorimotor
cortex. It also extends to the SM A, premotor cortex, cingulate cortex, ipsilateral sensorimotor cortex, left inferior parietal
cortex (SI, SII, BA 40), bilateral insular cortex and parietooccipital sulci. The negative eigenimage (Fig. 5D) shows the
functional network involved in the matching task. The pattern
shows a greater co-activation of the right hemisphere. Its highest
loadings are in the right inferior and middle temporal gyri and in
the medial part of the superior frontal gyrus (BA 8/9). All the
prefrontal cortex and the right IPS seems to be involved in the
matching task.
Discussion
The aim of this study was to determine the neural substrate of
object processing according to the task in which the subject is
involved. Cortical rCBF was measured during a task involving
visuomotor transformation (grasping objects) and a task
involving perceptual visual judgement (matching object shapes
with each other). As the same set of stimuli was used in both
conditions, only the task differed. These two tasks both require
analysis of object properties, but for different purposes. In the
case of grasping, object size, volume and shape are analyzed for
programming and executing the prehension movement (see
Jeannerod, 1981). In the case of matching, the analysis of object
shape is made for the purpose of memorizing and comparing.
The results of this experiment were twofold. First, the activation
of some cortical areas was task-specific: whereas the inferior
parietal cortex in the contralateral hemisphere was involved in
the network performing the visuomotor transformation,
perceptual matching involved the temporal cortex on the right
side. Second, and most importantly, a limited cortical area in the
right intraparietal sulcus was common to both tasks.
Cerebral Cortex Jan/Feb 1997, V 7 N 1 81
Figure 4. Cerebral regions activated during the grasping task compared with the matching task. All conventions are as for Figure 2.
Temporal Cortex
The significant rCBF increase in the temporal cortex during the
matching task was limited to a restricted region in the Brodmann
area 37 (Fig. 2). This local increase in blood f low, however, was
included in a widespread functional network, as shown in Figure
5D. This network extends over the occipito-temporal junction,
the middle and inferior temporal gyri, a large fraction of the
prefrontal lobe and the posterior parietal cortex. Similar results
(except for the parietal activation discussed below) were
obtained in previous studies dealing with visual processing of
objects or faces both in humans (Haxby et al., 1991; Sergent et
82 Cerebral Regions Involvement in Grasping and Shape Matching • Faillenot et al.
al., 1992; Kosslyn et al., 1994) and non-human primates
(Mishkin et al., 1983; Bachevalier and Mishkin, 1986; Tanaka,
1993). A recent study on shape matching using unknown
abstract objects found a restricted activation in the posterior
inferior temporal gyrus (BA 37) (Smith et al., 1995), i.e. lateral
with respect to the area activated during the processing of
unknown faces which involved mainly the fusiform gyrus.
The fact that the right hemisphere was more involved than the
left one in our matching task is also consistent with previous
reports by Horwitz et al. (1992) and by McIntosh et al. (1994). In
these studies, the object vision condition was similar to ours: it
Figure 5. Results of principal components extraction. (A) The SVD analysis showed two factors. Factor 1 accounts for 77.5% of variance of all the data. (B) The main factor is related
to the difference between the grasping task (G) and the matching task (M). (C) The positive eigenimage corresponds mainly to the functional network involved in the grasping task.
(D) The negative image shows the network involved in the matching task. The eigenimages are displayed as a maximum intensity projection in standard SPM format. The color scale
is arbitrary and each image is scaled to its maximum. The numbers indicate the size in mm of the stereotactic box. R indicates the right hemisphere. VAC and VPC are the verticals
traversing the anterior and posterior commissures.
required that the attributes of objects (faces) be analyzed for the
purpose of comparison. Sergent et al. (1992) also reported an
activation at the right occipito-temporal junction (BA 19, 37)
during face identification, whereas the left occipito-temporal
cortex (BA 19, 20, 21) was activated during object categorization. This observation is concordant with behavioral data
suggesting the existence of two independent visual-form systems
(e.g. Marsolek, 1995). One of them, which operates in the right
hemisphere, stores local details for distinguishing specific
instances of a given form. The other one, in the left hemisphere,
stores invariant information for discriminating different form
categories. Thus, the activation of cortical areas in the right
hemisphere during our matching task would be consistent with
this model.
Anterior Parietal Cortex
During the grasping–matching subtraction, areas involved in
somatosensory processing became visible (Fig. 4). This pattern
of activation also appears on the positive eigenimage of
correlated activity analysis (Fig. 5C). In the parietal lobe, the
rCBF increased in primary somatosensory areas (BA 1, 2 and 3)
at the level of the hand representation (Z > +40 mm; see Matelli
Cerebral Cortex Jan/Feb 1997, V 7 N 1 83
et al., 1993; Grafton et al., 1994; Sadato et al., 1996). A distinct
focus of activity was located in the left inferior postcentral
sulcus (BA 2/40) and extended towards the secondary sensory
area (SII) and the posterior insular cortex. The latter two regions,
which are anatomically linked, have often been shown to be
coactivated (e.g. Seitz and Roland, 1992; Burton et al., 1993;
Stephan et al., 1995). Thus the parietal operculum was involved
in both the pointing and the grasping tasks, but the ventral part
of the postcentral sulcus was activated by the grasping task only
(see also Grafton et al., 1996).
The finding of a specific activation in the post-central sulcus
during the grasping task is surprising. Based on the localization
of lesions affecting visuomotor transformation (Perenin and
Vighetto, 1988), we had predicted that grasping would involve a
more posterior region in the parietal lobe. Instead, we found
activation in an anterior region (BA 2/40, Z < +32 mm) when the
pointing task was subtracted from grasping. Activation of the
inferior part of the postcentral sulcus was reported in a number
of different tasks, such as stimulation by moving visual stimuli
(Dupont et al., 1994), tactile discrimination (O’Sullivan et al.,
1994) and mental imagery of grasping movements (Decety et al.,
1994). In addition, Grafton et al. (1996) found an area in the
lateral parietal operculum involved in the grasping task but not
in the pointing task. Its anatomical location as well as its
Talairach coordinates closely correspond to the ventral part of
the area activated in the present study. These results are
consistent with Deiber et al. (1996), who showed that the
anterior parietal cortex (BA 40) close to the postcentral sulcus
plays a critical role in the use of visual instructions for motor
preparation. The Talairach coordinates of this area suggest that it
overlaps the dorsal part of the area involved in the present
grasping task.
Considering its function and its anatomical location, the most
anterior part of the human IPL bears some homology to the
inferior parietal area 7b in the monkey (Hyvärinen, 1981). This
latter area, which is connected to other parietal areas such as SII,
A IP and VIP and to the ventral premotor cortex, is part of a
network for producing motor responses to visual and
somatosensory stimuli (Graziano et al., 1994; Graziano and
Gross, 1993).
Occipito-parietal Cortex
The shape matching task activated a region in the right
occipito-parietal cortex which was found not to be concerned in
either the grasping or the pointing task. This region is at the
border between Brodmann areas 19 and 7. Haxby et al. (1994),
who found that this area was activated in a location matching
task, concluded that it was involved in the processing of spatial
vision. This can hardly be the case in the present experiment,
since our two tasks both carried spatial location analysis.
Another possible interpretation (Kosslyn et al., 1994, 1995) is
that this region is active during object identification tasks
because it is involved in shifting attention to the location of an
expected object part or property. Since the stimuli used in our
study were abstract and unknown, devoid of semantic or
functional cues, shape comparison depended exclusively on
perceptual information. It is likely that in these conditions, the
attentional component was more important in the matching task
than in the other tasks (as the subjects themselves reported).
Because the matching task relied, at least in part, on localizing
the area of the object where the critical properties were located,
subjects had to orient their local attention. Whatever the
mechanism involved (perceptual or attentional), it is logical to
84 Cerebral Regions Involvement in Grasping and Shape Matching • Faillenot et al.
submit that this part of the occipito-parietal cortex is involved in
the local analysis of object properties.
One cannot exclude that the predominent right hemisphere
activation in the matching condition was due, at least partly, to
the subtraction methodology: when the grasping (or the
pointing) condition was subtracted from matching, a predominant activation in the left fronto-parietal cortex was removed.
Intraparietal Sulcus
Both the grasping and, to a greater extent, the matching tasks
were associated with an rCBF increase in the IPS (BA 40/7). This
area is thus clearly associated with the visual analysis of objects.
Indeed, both tasks required visual processing in order to analyze
the size, volume and orientation of the object irrespective of its
spatial location. During the grasping task, the object size and
volume determined the grip size and the adequate number of
fingers that were necessary to grasp it; during the matching task
size and volume had to be processed in order to extract the
invariant shape properties that were used for the comparison.
A possible interpretation is that this area in the IPS plays a role
in integrating the three-dimensional properties of objects, based
on the finding that visual IPS neurons in the monkey respond to
three-dimensional object properties (axis or surface orientation,
thickness, shape) from binocular disparity signals (Ohtsuka et
al., 1995). Further studies are needed to determine the specific
role of IPS in visual object analysis.
Conclusion
The present results do not fully support the dissociation
between visual pathways specialized for ‘perception’ (the
occipito-temporal pathway) and for ‘action’ (the occipitoparietal
pathway) (e.g. Goodale and Milner, 1992), which would have
predicted that a specific region in the posterior parietal cortex is
distinctly involved in grasping as compared with matching. This
region was not found in the present study. Rather, an area in the
IPS, common to the two tasks, was activated. This finding
underlines that recognition and grasping are not independent
processes using strictly parallel pathways. Rather, our results
support the idea that object-oriented action and shape
comparison both involve a form of visual object analysis which
is processed by the same posterior parietal area. This mechanism
would be distinct from the analysis of shape mediated by the
temporal cortex, and its exclusion by a lesion would not prevent
object recognition.
Notes
The authors are grateful to M. T. Perenin and Y. Rossetti for their advice
on experimental design. I.F. was supported by a grant from the Human
Science Frontier Program. I.T. was supported by the Fondation Fyssen
(Paris).
Correspondence should be adressed to Dr Jean Decety, Vision et
Motricité, INSERM U.94, 16 Av du Doyen Lépine, F-69500 Bron, France.
References
Bachevalier J, Mishkin M (1986) Visual recognition impairment follows
ventromedial prefrontal lesions in monkeys. Behav Brain Res
20:249–261.
Boussaoud D, Ungerleider LG, Desimone R (1990) Pathways for motion
analysis: cortical connections of the medial superior temporal and
fundus of the superior temporal visual areas in the macaque. J Comp
Neurol 296:462–495.
Burton H, Videen TO, Raichle ME (1993) Tactile-vibration-activated foci
in insular and parietal-opercular cortex with PET: mapping the second
somatosensory area in humans. Somatosens Motor Res 10:297–308.
Decety J, Perani D, Jeannerod M, Bettinardi V, Tadary B, Mazziotta JC,
Woods R, Fazio F (1994) Mapping motor representations with
positron emission tomography. Nature 371:600–602.
Deiber MP, Ibanez V, Sadato N, Hallett M (1996) Cerebral structures
participating in motor preparation in humans: a positron emission
tomography study. J Neurophysiol 75:233–247.
Dupont P, Orban GA, Bruyn Bd, Verbruggen A, Mortelmans L (1994)
Many areas in the human brain respond to visual motion. J
Neurophysiol 72:1420–1424.
Farah MJ (1992) Agnosia. Curr Opin Neurobiol 2:162–164.
Faugier-Grimaud S, Frenois C, Stein DG (1978) Effects of posterior
parietal lesions on visually guided behavior in monkeys.
Neuropsychologia 16:151–168.
Friston KJ (1994) Functional and effective connectivity in neuroimaging:
a synthesis. Hum Brain Mapping 2:56–78.
Friston KJ, Passingham RE, Nutt JG, Heather JD, Sawle GV, Frackowiak
RSJ (1989) Localisation in PET images: direct fitting of the intercommissural (AC–PC) line. J Cereb Blood Flow Metab 9:690–695.
Gallese V, Murata A, Kaseda M, Niki N, Sakata H (1994) Deficit of hand
preshaping after muscimol injection in monkey parietal cortex.
NeuroReport 5:1525–1529.
Goodale MA, Milner AD (1992) Separate visual pathways for perception
and action. Trends Neurosci 15:20–25.
Goodale MA, Milner AD, Jacobson LS, Carey DP (1991) A neurological
dissociation between perceiving objects and grasping them. Nature
349:154–156.
Grafton ST, Fagg AH, Woods RP, Arbib MA (1996) Functional anatomy of
pointing and grasping in humans. Cereb Cortex 6:226–237.
Grafton ST, Wood RP, Tyszka M (1994) Functional imaging of procedural
motor learning: relating cerebral blood f low with individual subject
performance. Hum Brain Mapping 1:221–234.
Graziano MSA, Gross CG (1993) A bimodal map of space: somatosensory
receptive fields in the macaque putamen with corresponding visual
receptive fields. Exp Brain Res 97:96–109.
Graziano SA, Yap GS, Gross CG (1994) Coding of visual space by
premotor neurons. Science 266:1054–1056.
Gross CG, Rocha-Miranda CE, Bender DB (1972) Visual properties of
neurons in the inferotemporal cortex of the macaque. J Neurophysiol
35:96–111.
Haxby JV, Grady CL, Horwitz B, Ungerleider LG, Mishkin M, Carson RE,
Herscovitch P, Schapiro MB, Rapoport SI (1991) Dissociation of object
and spatial visual processing pathways in human extrastriate cortex.
Proc Natl Acad Sci USA 88:1621–1625.
Haxby JV, Horwitz B, Ungerleider LG, Maisog JM, Pietrini P, Grady CL
(1994) The functional organization of human extrastriate cortex: a
PET-rCBF study of selective attention to faces and locations. J Neurosci
14:6336–6353.
Horwitz B, Grady CL, Haxby JV, Schapiro MB, Rapoport SI, Ungerleider
LG, Mishkin M (1992) Functional associations among human posterior
extratriate brain regions during object and spatial vision. J Cogn
Neurosci 4:311–322.
Hy värinen J (1981) Regional distribution of functions in parietal
association area 7 of the monkey. Brain Res 206:287–303.
Jeannerod M (1981) Intersegmental coordination during reaching and
natural visual objects. In: Attention and performance (Long J,
Baddeley A, eds), pp 153–168. Hillsdale, NJ: Lawrence Erlbaum.
Jeannerod M (1986) The formation of finger grip during prehension, a
cortically mediated visuomotor pattern. Behav Brain Res 19:99–116.
Kosslyn SM, Alpert NM, Thompson WL, Rauch SL, Anderson AK (1994)
Identifying objects seen from different viewpoints: PET investigation.
Brain 117: 1055–1071.
Kosslyn SM, Alpert NM, Thompson WL (1995) Identifying objects at
different levels of hierarchy: a PET study. Hum Brain Mapping
3:107–132.
Marsolek CJ (1995) Abstract visual-form representations in the left
cerebral hemisphere. J Exp Psychol Hum Percept Perf 21:375–386.
Matelli M, R izzolatti G, Bettinardi V, Gilardi MC (1993) Activation of
precentral and mesial motor areas during the execution of elementary
proximal and distal arm movements: a PET study. NeuroReport
4:1295–1298.
Mazoyer B, Trebossen R, Shoukroun C, Verrey B, Syrota A, Vacher J,
Lemasson P (1990) Physical caracteristics of TTV03, a new hight
spatial resolution time-of-f light positron tomograph. IEEE Trans Nucl
Sci 37:778–782.
McIntosh AR, Grady CR, Ungerleider LG, Haxby JV, Rapoport SI, Horwitz
B (1994) Network analysis of cortical visual pathways mapped with
PET. J Neurosci 14:655–666.
Mishkin M, Ungerleider LG, Macko K A (1983) Object vision and spatial
vision: two cortical pathways. Trends Neurosci 6:414–417.
Morel A, Bullier J (1990) Anatomical segregation of two cortical visual
pathways in the macaque monkey. Vis Neurosci 4:555–578.
O’Sullivan BT, Roland PE, Kawashima R (1994) A PET study of
somatosensor y discrimination. Microgeometry versus macrogeometry. Eur J Neurosci 6:137–148.
Ohtsuka H, Tanaka Y, Kusunoki M, Sakata H (1995) Neurons in monkey
parietal association cortex sensitive to axis orientation. J Jpn
Ophthalmol Soc 99:59–67.
Oldfield RC (1971) The assessment and analysis of handedness: the
Edinburgh inventory. Neuropsychologia 9:97–113.
Perenin MT, Vighetto A (1988) Optic ataxia: a specific disruption in
visuomotor mechanisms. Brain 111:643–674.
Perrett DI, Rolls ET, Caan W (1982) Visual neurons responsive to faces in
the monkey temporal cortex. Exp Brain Res 47:329–342.
Perrett DI, Mistlin AJ, Chitty AJ (1987) Visual cells responsive to faces.
Trends Neurosci 10:358–364.
Pohl W (1973) Dissociation of spatial discrimination deficits following
temporal and parietal lesions in monkeys. J Comp Physiol Psychol
82:227–239.
Raichle ME (1987) Circulatory and metabolic correlates of brain function
in normal humans. In: Handbook of physiology, the nervous system,
part 2 (Mountcastle VB, Plum F, Geiger SR, eds), pp 643–674.
Bethesda, MD: Waverly Press.
Sadato N, Ibanez V, Deiber M-P, Campbell G, Leonardo M, Hallett M
(1996) Frequency-dependent changes of regional cerebral blood f low
during finger movements. J Cereb Blood Flow Metab 16:23–33.
Sakata H, Taira M (1994) Parietal control of hand action. Curr Opin
Neurobiol 4:847–856.
Seitz RJ, Roland PE (1992) Vibratory stimulation increases and decreases
the regional cerebral blood f low and oxidative metabolism: a PET
study. Act Neurol Scand 86:60–67.
Sergent J, Ohta S, Macdonald B (1992) Functional neuroanatomy of face
and object processing — a PET study. Brain 115:15–36.
Smith EE, Jonides J, Koeppe R A, Awh E, Schumacher EH, Minoshima S
(1995) Spatial versus object working memor y: PET investigation.
J Cogn Neurosci 7:337–356.
Stephan KM, Fink GR, Passingham RE, Silbersweig D, Ceballos-Baumann
AO, Frith CD, Frackowiack RSJ (1995) Functional anatomy of the
mental representation of upper extremity movements in healthy
subjects. J Neurophysiol 73:373–386.
Taira M, Mine S, Georgopoulos AP, Murata A, Sakata H (1990) Parietal
cortex neurons related to the visual guidance of hand movement. Exp
Brain Res 83:29–36.
Talairach J, Tournoux P (1988) Co-planar stereotaxic atlas of the human
brain. George W. Thieme: Stuttgart.
Tanaka K (1993) Neuronal mechanisms of objets recognition. Science
262:685–688.
Tanaka K, Saito H, Fukada Y, Moriya M (1990) Integration of form, texture
and color information in the inferotemporal cortex of the macaque.
In: Vision memory and the temporal lobe (Iwai H, Misitkin M, eds), pp
101–109. New York: Elsevier.
Woods RP, Mazziotta JC, Cherry SR (1993) MRI–PET registration with
automated algorithm. J Comput Assist Tomogr 17:536–546.
Cerebral Cortex Jan/Feb 1997, V 7 N 1 85