1

Title: Discrete parieto-frontal functional connectivity related to grasping

Authors:

Noriaki Hattori, Hiroshi Shibasaki, Lewis Wheaton, Tao Wu, Masao

Matsuhashi, Mark Hallett

Affiliation:

Human Motor Control Section, Medical Neurology Branch, National Institute of

Neurological Disorders and Stroke, National Institutes of Health, Bethesda,

Maryland, USA

Running head: Parieto-frontal network related to grasping

Full address, e-mail address, phone and fax number of corresponding author:

Hiroshi Shibasaki, Takeda General Hospital, Ishida, Fushimi-ku, Kyoto,

601-1495 Japan, shib@kuhp.kyoto-u.ac.jp, phone: +81-75-572-6468. fax:

+81-75-571-8877

Articles in PresS. J Neurophysiol (December 24, 2008). doi:10.1152/jn.90249.2008

Copyright © 2008 by the American Physiological Society.

2

Abstract

The human inferior parietal lobule (IPL) is known to have neuronal connections

with the frontal lobe, and these connections have been shown to be associated

with sensorimotor integration to perform various types of movement such as

grasping. The function of these anatomical connections has not been fully

investigated. We studied the judgment of graspability of objects in an

event-related functional MRI study in healthy subjects, and found activation in

two different regions within IPL; one in the left dorsal IPL extending to the

intraparietal sulcus and the other in the left ventral IPL. The former region

was activated only in the judgment of graspable objects while the latter was

activated in the judgment of both graspable and non-graspable objects although

the activation was greater for the graspable objects. Psychophysiological

interaction analysis showed that these regions had similar, but discrete

functional connectivity to the lateral and medial frontal cortices. In relation to

this particular task, the left dorsal IPL had functional connectivity to the left

ventral premotor cortex, supplementary motor area (SMA) and right cerebellar

3

cortex, whereas the left ventral IPL had functional connectivity to the left

dorsolateral prefrontal cortex and pre-SMA. These findings suggest that the

connection from the left dorsal IPL is associated specifically with automatic flow

of information about grasping behavior. By contrast, the connection from the

left ventral IPL might be related to motor imagination or enhanced external

attention to the presented stimuli.

4

Introduction

Grasping is one of the most fundamental behaviors for human and non-human

primates. In order to grasp an object, we have to shape our hands and fingers

precisely before touching the object. This process includes transforming

intrinsic properties of the object into motor actions. The results of single unit

recording studies on monkeys have indicated that neurons in the anterior

intraparietal area (AIP) and caudal ventral premotor cortex (F5) are activated

when monkeys grasp visually presented objects (for review: Rizzolatti and

Luppino (2001) and Shikata et al. (2003)).

Human clinical studies have shown that damage to the posterior parietal area

causes impairment of grasping behavior. Binkofski et al. (1998) demonstrated

that patients with lesions involving the anterior lateral bank of the

intraparietal sulcus showed selective deficits in the coordination of finger

movements required for object grasping, whereas patients with parietal lesions

sparing this region showed intact grasping behavior. Jeannerod et al. (1994)

5

reported a patient with bilateral posterior parietal lesions who presented

bilateral deficit in grasping objects without deficit in reaching. Furthermore,

neuroimaging studies using positron emission tomography and functional

magnetic resonance imaging (fMRI) have been applied to the execution,

imagination and observation of grasping. Multiple brain areas including

contralateral primary motor cortex, bilateral premotor cortex (PM),

supplementary motor area (SMA) and posterior parietal cortex are activated

during grasping and reaching movement (Binkofski et al. 1998; Grafton et al.

1996b). Furthermore, the mere viewing of graspable objects or tools activates

several brain areas, such as the left PM and posterior parietal cortex (Chao and

Martin 2000; Creem-Regehr and Lee 2005; Grafton et al. 1997; Grezes et al.

2003). These findings suggest that humans have neurons possessing similar

properties to those of the object-type neurons in AIP (Taira et al. 1990) and

canonical neurons in F5 (Murata et al. 1997; Rizzolatti et al. 1988) of monkeys.

Neuroimaging studies have been applied to map the human AIP (Culham et al.

2003; Grefkes and Fink 2005; Frey et al. 2005; Culham et al. 2006; Tunik et al.

2007), but its precise location and task-specific interaction with the ventral

6

premotor cortex (human F5) still remain to be explained.

In this study, we conducted an event-related fMRI study in which subjects were

asked to judge whether visually presented objects were graspable or not. We

hypothesized that the judgment is based on whether proper motor program of

grasping each object is successfully retrieved or not, and that human AIP is

related to this process. We further hypothesized that the inherent functional

connectivity between human AIP and F5 could be modulated depending on the

judgment, and this modulation could be shown by psychophysiological

interaction (PPI) analysis.

Materials and Methods

Subjects

Seventeen healthy subjects (nine females, eight males) aged 21 to 57 years

(mean age 33 years) were recruited from the National Institutes of Health

7

database of normal volunteers. All subjects were right-handed according to

the Edinburgh Handedness Inventory (Oldfield 1971), and neurologically

normal. Their visual acuity was either normal or corrected by contact lenses.

Informed verbal and written consents for this protocol, which was approved by

the National Institute of Neurological Disorders and Stroke Institutional

Review Board, were obtained from all the subjects.

Stimulus

Visual stimuli were prepared from a set of 260 line-drawn pictures, which were

standardized by Snodgrass and Vanderwart in terms of four variables of central

relevance to memory and cognitive processing: name agreement, familiarity,

visual complexity, and image agreement (Snodgrass and Vanderwart 1980).

We selected 50 pictures from this list, so that about half of them were expected

to be judged as graspable objects, and the remaining half as non-graspable.

Graspable objects consisted of pictures of tools, utensils, foods, and other

relatively small objects, which can be easily grasped by one hand.

Non-graspable objects were those not graspable by one hand because of the

8

configuration, texture, size, or the purpose of use (Table 1).

Although tools were the most common category among graspable objects, we did

not restrict the graspable objects to tools. This is because ‘affordances’ (Arbib

1997) of tools are supposed to be strongly connected to their usage, and

perception of tools has been reported to directly activate the brain areas

associated with motor performance (Chao and Martin 2000; Creem-Regehr and

Lee 2005; Grafton et al. 1997). In particular, the type of evoked movement by

tools is not necessarily grasping, but rather a hand movement specific for the

use of each tool.

Visual stimuli were generated by a personal computer using Presentation

software (Neurobehavioral Systems, Inc., Albany, CA). Subjects lay supine on

the scanner bed and looked at a mirror fixed on the head coil, allowing them to

see an opaque screen located at their feet, on which visual stimuli were rear

projected. The pictures were presented in black on white background at the

visual angle of approximately 5 degrees in both the horizontal and vertical

9

directions. To control the size effect, the size of each picture was proportional

to the size of the original pictures reported by Snodgrass and Vanderwart (1980),

rather than to the actual object size.

Experimental procedure

The experiment consisted of two conditions; ‘judgment’ and ‘resting’. In the

judgment condition, the pictures of graspable and non-graspable objects were

presented in random sequence at a fixed interval of 6.3 s. Each picture was

shown for 600 ms. The subjects were asked to judge whether each presented

object was graspable by one hand or not. The left or right hand was not

specified to the subjects in order to focus on the cognitive aspect of grasping and

minimizing the possible motor effect. However, since the subjects were

all right handed, they might be expected to have more

experience grasping these objects with the right hand. They were instructed to

make judgment internally, and to avoid execution of any overt movement. In

each judgment condition, pictures of 10 objects were presented. In the resting

condition, the subjects were asked to relax and just fixate on black crosshairs

10

shown in the center of the screen. Each judgment condition lasted 57.3 s, and

each resting condition 32.7 s. Each subject had a single scanning session

which consisted of alternate presentation of the judgment and resting

conditions 5 times each, and the session always started with the resting

condition. After the image acquisition, the subjects were shown the same set of

pictures on a computer display, and reported to the experimenter their

judgment of each picture as to whether graspable or non-graspable. For the

analysis of fMRI data, categorization of each picture as to graspable or

non-graspable was made according to the subject’s own judgment instead of the

judgment prepared by the experimenter in advance.

Functional MRI scanning

Blood oxygenation level-dependent (BOLD) contrast functional images (Ogawa

et al. 1993) were obtained using two equivalent whole-body 3.0 Tesla MRI

systems (Signa, General Electric, Milwaukee, WI) equipped with gradients

capable of 40 mT/m amplitude and 150 T · m-1 · s-1 slew rate. Images were

acquired by using gradient-echo echo-planar imaging (EPI) (TR/TE = 2500

11

ms/25ms, FA = 90°, slice thickness/gap = 5/1mm, FOV = 22 X 22 cm2, matrix =

64 X 64), and twenty-two slices were obtained. Head motion was reduced by a

vacuum pillow placed around the subject's head. The time series began with

dummy gradients and RF pulses corresponding to the first four images to allow

brain tissue to reach steady-state magnetization, and 180 volumes of images in

total were acquired.

Image processing

Off-line data processing and analysis were performed using the Statistical

Parametric Mapping Software (SPM2; Wellcome Department of Imaging

Neuroscience, London, UK) implemented within Matlab 7 (MathWorks,

Sherborn, MA). Image volumes were corrected for slice timing skew using

temporal sinc interpolation and realigned to the first acquisition using

rigid-body transformation. The mean image of the realigned images was

spatially normalized to the standard EPI template proposed as default in SPM2.

This SPM template is in Montreal Neurological Institute (MNI) space (Montreal

Neurological Institute, http://www.bic.mni.mcgill.ca) and approximates the

12

standard stereotaxic space of Talairach and Tournoux (Talairach and Tournoux

1988). Linear and nonlinear deformation parameters estimated during this

step were then subsequently applied to all the realigned EPI volumes of the

corresponding time series. Spatially normalized images were smoothed using a

6 mm full-width at half-maximum Gaussian kernel.

Image analysis

Realigned, spatially normalized, and smoothed T2*-weighted EPI images were

analyzed using SPM2 in the framework of the general linear model (Friston et

al. 1995). Both first- and second-level analyses were performed. In the

first-level, fixed-effects analysis, data were analyzed for each individual subject

separately on a voxel-by-voxel basis using the principles of the general linear

model extended to allow the analysis of fMRI data as a time series.

Event-related responses to the onset of the presentation of the object pictures

were examined, with the model including regressors for 1) judgment of the

presented object as graspable (G-judgment condition) and 2) judgment of the

presented object as non-graspable (N-judgment condition) depending on the

13

judgment made by each subject. Regressors were created based on the

canonical hemodynamic response function, and temporal derivatives of the

hemodynamic response function were included as additional regressors. All

data were high-pass filtered with a cutoff frequency of 1/180 Hz to remove

low-frequency signal drifts, and a correction for temporal autocorrelation was

applied. Contrasts representing the effect of the G-judgment and N-judgment

conditions compared with baseline (the resting condition) were defined and

contrast images were calculated respectively. The calculated contrast images

from each subject were used for the second level analysis.

Random-effects analysis took into account both of the two sources of variability;

within-subject and between-subject variability. To make inferences about the

population from which the subjects were drawn, individual contrast images

from the first-level analysis were entered in the second-level, random-effects

analysis (Penny and Holmes 2003). A conjunction analysis (Friston et al. 2005;

Nichols et al. 2005) was performed to identify the positive changes in BOLD

signal intensity which were commonly seen in the G-judgment and N-judgment

14

conditions by using contrast images of each condition compared with the resting

condition and using analysis of variance (ANOVA) model in SPM2.

Significance level was set at p < 0.05, voxel-level corrected for multiple

comparisons using the false discovery rate (Genovese et al. 2002). Next, to

identify the difference in the activation between the two conditions, which was

expected to be smaller than the activation commonly identified in the two

conditions, we introduced a different analysis method. Although we

hypothesized that the left inferior parietal lobule (IPL) and/or PM would be

activated in this comparison, there might be other locations showing

significantly different activation between these two conditions, such as the

temporal lobe in relation to semantic memory of objects. For this reason, first

we performed the statistical analysis over the whole brain rather than

restricting the analysis to selected regions of interest (ROI) with a priori

hypothesis, which potentially might miss the activation outside the ROI. A

paired t test model was applied to the contrast images representing the

differences in the activation of the two judgment conditions of individual

subjects. Significance level was set at p < 0.001, voxel-level without correction

15

for multiple comparisons, and then only clusters with p < 0.05, cluster-level

with correction for multiple comparisons, were identified as significant. The

extent threshold was set to 10 voxels for the second-level analysis.

Next, supplementary analysis was performed for the contrasts among different

functional categories determined by the result of the judgment and inherent

characteristics of the objects. The objects most subjects judged as graspable

and those non-graspable were further categorized into subgroups (Table 1), and

brain activations were compared among these functional categories. The aim

of this analysis was to elucidate if there was any difference in the activation

among different object properties. The analysis method and statistical

threshold were the same as for the analysis between G-judgment and

N-judgment conditions.

ROI analysis was performed in order to directly compare the activation

associated with retrieval of the motor program of grasping in IPL and PM.

Local percent signal changes associated with each judgment condition compared

16

with the baseline signal intensity were calculated at each voxel in each

individual subject, and the percent signal changes associated with the

N-judgment condition were subtracted from those associated with the

G-judgment condition. Second, voxels were chosen as “seeds” for searching

individually optimized ROI in IPL and PM from the peaks of the clusters

detected by the above-described method with the contrast of the G-judgment

condition minus the N-judgment condition. The voxels were thresholded at

significance level of p < 0.001 without correction for multiple comparisons.

Third, in order to treat the variability of anatomical location of activated

regions among the subjects, spheres with the center positioned on these voxels

with a radius of 10 mm were located on the image of the subtracted percent

signal change of the individual subject, and voxels with maximum percent

signal change within these spheres were chosen as the center of ROI for the ROI

analysis. Then, spherical ROIs with a radius of 6 mm were located in the

image of the subtracted percent signal change of the individual subject. An

SPM tool box, MarsBaR (Brett et al. 2002) was used to locate ROIs. The mean

value of the percent signal change within the ROI of the individual subject was

17

calculated and these mean values were analyzed by repeated measures ANOVA,

and if the within-subjects effects were found significant, then post-hoc pairwise

comparisons were performed. Significance level was set at p < 0.05, and

SPSS12 software (SPSS Inc., Chicago, Illinois) was used for repeated measures

ANOVA.

Psychophysiological interaction (PPI) analysis

Finally, in order to show judgment-related functional connectivity in these IPL

and PM ROIs, PPI analysis was performed. PPI analysis employs a first

regressor representing the (deconvolved) activation time course in the index

ROI (the physiological variable), a second regressor representing psychological

variable (G versus N judgment) and a third regressor representing the

interaction of these two factors (psychophysiological interaction). In the PPI

analysis, only the third regressor is used to create a contrast of interest in the

statistical analysis. Since the interaction is expressed, not at the level of

hemodynamic responses, but at a neuronal level, the BOLD signal in the index

ROI was deconvolved with the canonical hemodynamic response function to

18

calculate the time course (Gitelman et al. 2003). The ROI extracted in the

above ROI analysis was again used for the PPI analysis for each subject as an

index ROI, and the brain areas which showed significantly positive modulation

of the functional connectivity in G-judgment compared with N-judgment were

calculated by using the PPI function implemented in SPM2. The calculated

contrast images from each subject were then used for the second level,

random-effects analysis. Significance level was set at p < 0.01, voxel-level

without correction for multiple comparisons, and only clusters with p < 0.05,

cluster-level with correction for multiple comparisons, were identified as

significant in the PPI analysis. This correction was performed in the voxels

which were activated in the G-judgment condition compared with the resting

state (uncorrected p <0.05). The extent threshold was set to 10 voxels for the

second-level analysis.

Anatomical localization was performed by superimposing the activation map

calculated in the second-level analysis on the mean spatially normalized EPI

images from all subjects as well as the MNI template. This approach takes

19

into account the variance of brain structures between subjects under

investigation and spatial distortions inherent to EPI images. Anatomical

labeling was performed with the aid of the atlas by Duvernoy (1999). As this

approach has some limitations, the designation of the anatomical structures is

tentative rather than definitive (Garraux et al. 2005). Locations of activated

areas for different conditions were displayed by superimposing them on the

MNI template for demonstration.

Results

Behavioral data

How each object was judged by the subjects is shown in Table 1 as the ratio of

the subjects who judged “graspable” among all the 17 subjects. As we had

expected, the judgment was consistent among subjects for the majority of

objects, but some objects were judged differently among subjects. Among the

50 objects presented in the experiment, 32 objects (64%) were judged in

20

concordance by all the subjects (the ratio of the judgment graspable 0 or 1 in

Table 1), and 6 objects (12%) were judged concordantly by less than 15 subjects

(the ratio between 0.18 and 0.82 in Table 1). In terms of individual subjects, the

subjects judged 27.1 out of 50 objects (mean, S.D. 2.1) as graspable.

As for the subcategories of objects, among 24 objects judged graspable by at

least 16 of 17 subjects, 10 objects were categorized into food or objects graspable

by one hand or placed in mouth (G-F) and 14 objects into the graspable small

manufactured or round objects (G-SO). Among 20 objects judged as

non-graspable by at least 15 of 17 subjects, 11 objects were categorized into the

non-graspable animals (NG-A) and 9 objects into the non-graspable large and/or

immobile objects (NG-LO). The remaining 6 objects included those objects that

might have graspable parts in them such as accordion, chair and car, relatively

small animals like cat and chicken, and the picture of an eye that could be

graspable if removed from the body. The last group was called the objects with

ambiguous graspability (Amb).

21

fMRI results

Common activation for graspable and non-graspable objects

Conjunction analysis by using contrast images of the G-judgment and

N-judgment conditions compared with the resting condition showed multiple

brain areas significantly activated including bilateral dorsal PM (PMd) and

ventral PM (PMv), left primary motor cortex, bilateral dorsolateral prefrontal

cortex (DLPF), inferior frontal gyrus, superior parietal lobule (SPL), IPL,

superior, middle and inferior temporal gyrus, parahippocampal gyrus, middle

occipital gyrus, cuneus, lingual gyrus and fusiform gyrus. In the midline

structures, peaks with significant activation were found in bilateral pre-SMA,

left SMA, bilateral anterior cingulate gyrus and left posterior cingulate gyrus.

In the subcortical areas, significant activation was detected in the lentiform

nucleus, thalamus and cerebellum, all bilaterally (Figure 1, Table 2).

Differential activation between graspable and non-graspable objects

Figure 2 shows the brain areas more activated in the G-judgment condition

compared with the N-judgment condition. The images were thresholded at p <

22

0.001 (uncorrected). Among these areas, the cluster level analysis showed

significant activation at two clusters in the left parietal lobe (p < 0.05, corrected)

(Table 3). Both clusters were located in the left IPL [Brodmann area (BA) 40].

The ventral IPL cluster was located in the inferior and anterior part of the IPL,

corresponding to a part of the supramarginal gyrus (Figure 2; top, large arrow

1), and the dorsal IPL cluster extended superiorly to the intraparietal sulcus

and anteriorly to the postcentral gyrus (BA 2) (Figure 2; top, large arrow 2). In

order to elucidate the characteristics of activation of these peaks to each

judgment condition, we performed a second-level analysis by using contrast

images of each of the G-judgment and N-judgment conditions compared with

the resting condition and using one-sample t test. The peak of the ventral IPL

cluster was significantly activated in both the G-judgment and N-judgment

conditions compared with the resting condition. In contrast, the peak of the

dorsal IPL cluster was activated only in the G-judgment condition (p < 0.05,

corrected). Parameter estimates of these peaks are presented in the bottom of

Figure 2. No cluster was found where activation in the N-judgment condition

was greater than in the G-judgment condition.

23

Differential activation among different functional categories

In the comparison among different functional categories (both within the same

graspability group and across the different graspability groups), bilateral

fusiform gyri were more activated in NG-LO compared with NG-A. The left

inferior temporal gyrus was more activated in G-SO compared with NG-LO.

Bilateral fusiform gyri and the left middle occipital gyrus were more activated

in NG-LO compared with G-F. Outside of the ventral visual pathway, only the

left IPL was more activated in G-F compared with NG-LO. This cluster mostly

overlapped with the ventral IPL cluster found with the contrast between

G-judgment and N-judgment conditions (Table 4).

ROI analysis

In addition to the peaks of the two clusters in the left IPL which were

significantly more activated in the G-judgment compared with N-judgment

condition at the cluster-level second level analysis (Figure 2, Table 3), a peak of

the cluster in the left PMv (coordinates in mm in MNI space: -52, 6, 24, Figure 2,

24

small arrow) was activated only with uncorrected p of 0.001 at the second level

analysis (T = 4.25, Z = 3.42), and was also chosen as a seed for positioning

individually optimized ROIs. Mauchly’s test of sphericity showed significant

within-subject effect (p = 0.048), and Greenhouse-Geisser correction was

applied to the repeated measures ANOVA. Test of within-subjects effects

demonstrated significant difference in the percent signal changes among these

three ROIs (F (1.5/23.996)=10.093, p < 0.005). The pairwise comparison test

demonstrated that the percent signal change in the dorsal IPL ROI was

significantly larger than those of the PMv and ventral IPL ROIs (p < 0.005,

Figure 3).

PPI analysis

PPI analysis with respect to the index regions in the left dorsal and ventral IPL

ROIs showed significantly higher modulation in the G-judgment condition

compared with the N-judgment condition in the following areas (Figure 4, Table

5). The left ventral IPL ROI was connected to two clusters which were located

in the pre-SMA and left DLPF (Figure 4A), and the left dorsal IPL ROI was

25

connected to three clusters which were located in the left PMv, SMA and the

right cerebellum (Figure 4B). No cluster of significant PPI modulation was

found when the left PMv ROI was employed as an index region. .

Discussion

Unique features of the present study

In this study, we used event-related fMRI to show the brain areas that might be

related to the motor program of grasping and its functional connectivity. The

subjects were asked to make judgments whether each object, presented as a

line-drawn picture, was graspable by one hand or not. By taking advantage of

event-related fMRI in which individual models were made for the analysis

(Wagner et al. 1998), we analyzed the data according to the judgment of each

object by each individual subject. Since we intended to exclude brain

activation associated with confounding cognitive and motor processes, the

subjects were instructed to avoid making any actual movement such as

26

pantomiming the grasp. Also, no responses beyond the internal judgment,

such as pressing a response button, were required during the scan. Another

important point is that, as described in the Methods, we did not restrict the

graspable objects to tools.

Models for grasping movement

By referring to several models that have been proposed to illustrate the

mechanism of grasping movement based on the neurophysiological studies in

monkeys (Fagg and Arbib 1998; Jeannerod et al. 1995; Rizzolatti and Luppino

2001), we hypothesized that the following cognitive processes might be involved

in the present experimental paradigm. Both G-judgment and N-judgment are

expected to go through the same pathways at least in the initial stage of

information processing; the output from the primary visual cortex goes through

the dorsal visual system carrying spatial information and through the ventral

visual system for object recognition (Ungerleider and Mishkin 1982). Since the

line-drawings, not realistic images, of objects were used in the present

experiment, semantic memory of the object was most likely used to retrieve

27

physical properties of the objects including size and weight.

Next, the subjects make judgments about the graspability of the presented

objects. We expected that the judgment for most objects was done by referring

to the stored motor program of grasping. However, the strategy might differ

among subjects, and for some objects, the judgment might be done simply by

referring to the semantic memory adherent to the objects, and for other objects,

not only attempt at forming proper finger position for grasping, but imagination

of reaching, rotating objects and self rotation might be also involved.

Common activation observed in the judgment of graspable and non-graspable

objects

There were multiple brain areas commonly activated in the two judgment

conditions (Figure 1, Table 2). Cortical regions belonging to the ventral visual

system including fusiform and parahippocampal gyri were activated commonly

for both conditions. SPL, a representative area of the dorsal visual system,

was also activated bilaterally in both conditions. In these visual systems, we

28

found no difference in the activation between the two judgment conditions.

However, the analysis with respect to the functional subcategories of objects

showed differences in the ventral visual pathways: bilateral fusiform gyri

were more activated in NG-LO than NG-A, the left inferior temporal gyrus was

more activated in G-SO than NG-LO, bilateral fusiform gyri and the left middle

occipital gyrus were more activated in NG-LO compared with G-F (Table 4).

These findings agree with the results of previous studies indicating the

existence of object-specific regions in the ventral visual system (for review:

Grill-Spector (2003)).

Decety et al. (1994) demonstrated that, during imagination of grasping objects,

activation was observed in the bilateral precentral gyrus, left prefrontal areas,

left IPL, bilateral anterior cingulate cortex, bilateral caudate nucleus and left

cerebellum. Grafton et al. (1996a) demonstrated that imagined grasping

activated the left inferior and middle frontal gyrus, left caudal IPL, left rostral

SMA proper and left PMd. Imagination of reaching activated PMd, medial

intraparietal sulcus and SPL (Johnson et al. 2002). Multiple brain areas

29

including SPL, PM, SMA, somatosensory cortex, DLPF and extrastriate regions

were reportedly activated during mental rotation of objects or subjects’ hands

(Cohen et al. 1996; Kosslyn et al. 1998; Richter et al. 2000). Obviously, motor

imagination of reaching, rotation and grasping share multiple brain areas in

common. In the present study, several areas reportedly associated with motor

imagination were also activated including bilateral PMv, IPL, inferior frontal

gyrus, pre-SMA, prefrontal cortex and anterior cingulate gyrus.

Differences in the activation between judgment of graspable and non-graspable

objects

Two clusters in the left IPL were more activated in the G-judgment condition

compared with the N-judgment condition (Figure 2, Table 3). By contrast, no

cluster in PMv was significantly more activated in the G-judgment condition

compared with the N-judgment condition in the cluster level analysis. This

finding suggests that the representation of grasping movement, which was

operationally retrieved in the current experiment, is stored in the left IPL.

30

Regarding the differences among different functional categories, only one

cluster in the left ventral IPL was more activated in G-F compared with NG-LO.

Recent neurophysiological study on PF/PFG (named by Pandya and Seltzer

(1982), partially overlapping with the area 7b in the convexity of IPL) of

monkeys suggested that PF plays an important role in organizing eating

behavior, and PFG plays a role in linking hand and mouth motor acts (Rozzi et

al. 2008). The activation related to G-F in the ventral IPL may be related to

these characteristics of PF/PFG. Failure of finding different activation in

other parts of IPL among different functional categories might be due to the low

statistical power with the relatively small number of presented objects.

The peak of the dorsal IPL cluster was significantly activated only in the

G-judgment condition compared with the resting condition, whereas the peak of

the ventral IPL cluster was activated in both the G-judgment and N-judgment

conditions (Figure 2). The results of the ROI analysis, which directly

compared differences in the signal changes of ROIs between the two judgment

conditions, showed greater signal changes in the dorsal IPL than in the ventral

31

IPL or PMv (Figure 3). Specific activation in the dorsal IPL for the

G-judgment condition and larger activation compared with the ventral IPL

suggest that the dorsal IPL most likely stores the motor program of grasping.

Previous neuroimaging studies suggest that a region in the human AIP is

involved in visually-guided grasping (Culham et al. 2003; Grefkes and Fink

2005; Frey et al. 2005; Culham et al. 2006; Tunik et al. 2007) and the dorsal IPL

detected in the present study most likely corresponds to this region.

By contrast with the dorsal IPL, activation of the ventral IPL in both the G- and

N-judgment conditions and yet greater activation in the G-judgment condition

suggests that the ventral IPL is considered to be related to the cognitive

processes which are involved in both conditions but more in the G-judgment

condition. These cognitive processes likely include motor imagination of

rotation, reaching and grasping (Decety et al. 1994; Cohen et al. 1996; Grafton

et al. 1996a; Kosslyn et al. 1998; Richter et al. 2000; Johnson et al. 2002) and

attention to the external stimulation (Knudsen 2007). Lesions in the left

supramarginal gyrus are known to cause ideomotor apraxia, a failure to

32

implement the internal representation of a gesture into the appropriate motor

actions associated with difficulties in motor imagery and pantomime (Rizzolatti

and Matteli 2003; Jeannerod 2006; Wheaton and Hallett 2007). Comparing

patients with ideomotor apraxia due to frontal lesions and those due to parietal

lesions, Heilman et al. (1982) found impairment in distinguishing

well-performed from poorly performed movements only in those with parietal

lesions. It can be postulated that the activation in the ventral IPL is, at least

in part, related to motor imagery and action representation, not specific for

grasping. Interestingly, Fogassi et al. (2005) found that neurons in the

convexity of IPL (PF/PFG complex) in monkeys are active not only during

execution of hand actions, but also during the observation of similar actions

done by another individual (‘parietal mirror neuron’). Neural tracer studies

have shown that there are various cortical connections between PF/PFG and

frontal or premotor cortices in monkeys (Petrides and Pandya1984; Rozzi et al.

2006). Considering structural analogy between the two species, the ventral

IPL might be a human homologue of PF/PFG complex and thus might play roles

in action organization and understanding the partner’s intention.

33

Functional connectivity

The results of the PPI analysis support the above interpretation. In the

present study, the PPI was expected to identify brain areas in which the degree

of coupling with the index IPL regions is significantly modulated by the nature

of the judgment (G-judgment or N-judgment). The presence of a significant

judgment-specific change in coupling between the index IPL region and other

brain areas can be interpreted in two ways: either as a context specific change in

the contribution of the index region to the target regions or the modulation of

responses in the target regions to the psychological variable (judgment of the

graspability) by the contribution from the index region (Friston et al, 1997).

When the index regions were placed in the left dorsal and ventral IPL, the

target regions were found both in the lateral and medial frontal cortex (Fig. 4),

and those areas were clearly discrete depending on the index region (Fig. 5).

The results of the PPI analysis indicate that retrieval of the motor program of

grasping stored in the left dorsal IPL modulates its functional connectivity with

34

its downstream brain areas. The functional connectivity between the left

dorsal IPL and the left PMv is likely to correspond to the parieto-frontal

network for grasping in monkeys (AIP and F5 neurons), which supports our

interpretation based on the ROI analysis that the left dorsal IPL is the human

AIP. Since SMA and cerebellum are known as parts of the motor-associated

cortices, the preferential interaction between the left dorsal IPL and those two

regions probably reflects modulation of the network related to the motor

program for grasping.

Similarly to the left ventral IPL, DLPF and pre-SMA have been reported to be

associated with motor imagination of rotation, reaching and grasping (Decety et

al. 1994; Cohen et al. 1996; Grafton et al. 1996a; Kosslyn et al. 1998; Richter et

al. 2000; Johnson et al. 2002), and attention (Knudsen 2007). Therefore, it is

postulated that, in the present experimental paradigm, motor imagination

and/or attention to the aimed stimuli (graspable objects) might have modulated

functional connectivity between the left ventral IPL and these cortical regions.

35

A question arises as to whether the left dorsal IPL stores representations of only

grasping movements or other types of movement as well. In an fMRI study by

Boronat et al. (2005), subjects viewed pairs of pictures or words denoting

manipulable objects and determined whether the two objects in each pair were

manipulated similarly or not, or whether they served the same function or not.

Greater activation was seen in the left IPL bordering the intraparietal sulcus in

the former condition than in the latter. Valyear et al. (2007) performed a

detailed fMRI study to evaluate activations produced by naming tools, other

graspable objects and non-graspable objects. They showed higher activation

within a discrete region of the left anterior intraparietal cortex for tools than for

other graspable objects, but did not find any difference between graspable and

non-graspable objects. They also demonstrated that the area activated by

viewing tools had partial overlap with the region of AIP activated by grasping

objects in the same subjects. These findings indicate that the representation of

various movements is stored in AIP and its adjacent areas. There may also be

an effect of different types of cognitive tasks (judgment, naming and actual

movement), and identification of the precise location of motor representation of

36

various movements remains to be investigated.

Limitations of the present experiment

There are some limitations in the present study. First, we did not force the

subjects to use any particular cognitive process in the judgment of graspability.

In other words, the judgment totally depended on the subjects’ own strategy.

This is in fact a unique feature of this study, and at the same time this might

have caused some ambiguity in the interpretation of the results. In order to

avoid this ambiguity, for example, we could have instructed the subjects to

imagine the attempt to form hand and fingers to judge the graspability. Such

kind of intentional and active imagination of the movement could activate

several brain areas including PMv more strongly in the G-judgment condition.

This kind of imagination, however, would make the interpretation of the

subtraction data (G-judgment minus N-judgment) more complicated.

Additionally, in the present study, we asked the subjects to report their

response immediately after the experiment. Although most subjects reported

37

their responses with confidence, we could not exclude potential source of errors

due to the uncertainty of the memory.

An interesting finding was marked left-lateralization of the activation in the

G-judgment condition. In the present study, as described above, we aimed at

focusing on the subject’s naive cognitive process and did not specify which hand

to use for the judgment. Since all subjects were right-handed, it is conceivable

that they unconsciously used their right hands for the judgment. This might

have enhanced the activation in the left hemisphere. Some studies employed

naming graspable objects as a task, and it also showed lateralization of the

parietal activation to the left (Chao et al. 2000, Valyear et al. 2007). In these

studies, language process is also considered to have affected the hemisphere

difference. In order to answer the question about interhemispherical

difference in the role of human AIP, further studies with left handers and

specifying the hand to use are required.

Finally, as neuroanatomical studies of gross cortical shape (Ono et al. 1990) and

38

a recent computational histological study (Caspers et al. 2006) demonstrated,

there are considerable inter-subject variations in the cytoarchitecture in the

inferior parietal cortex. Despite the substantial variability in the intraparietal

sulcus topology, Frey et al. (2005) demonstrated that a contrast between pincer

grasping and reaching to complex shapes revealed activation foci at the junction

of the intraparietal sulcus and postcentral gyrus in all 14 individuals examined.

The spatial normalization technique based on the gross anatomical structure

that was used in the present study and in many of other functional imaging

studies may blur the shape of the activation in this particular cortical region.

We adjusted the location of the peak of activation for each subject to minimize

the effect of this inter-subject variation, but some ambiguity in terms of the

spatial localization might still remain.

Grants

The study was supported by the Intramural Research Program of NINDS.

39

References

Arbib MA. From visual affordances in monkey parietal cortex to

hippocampo-parietal interactions underlying rat navigation. Philos Trans

R Soc Lond B Biol Sci 352: 1429-1436, 1997.

Binkofski F, Dohle C, Posse S, Stephan KM, Hefter H, Seitz RJ, Freund HJ.

Human anterior intraparietal area subserves prehension: a combined

lesion and functional MRI activation study. Neurology 50: 1253-1259,

1998.

Boronat CB, Buxbaum LJ, Coslett HB, Tang K, Saffran EM, Kimberg DY, Detre

JA. Distinctions between manipulation and function knowledge of

objects: evidence from functional magnetic resonance imaging. Brain Res

Cogn Brain Res 23: 361-373, 2005.

Brett M, Anton J, Valabregue R, Poline J. Region of interest analysis using an

SPM toolbox. in Proceedings of the 8th International Conference on

Functional Mapping of the Human Brain, June 2-6, 2002, Sendai, Japan.

Neuroimage 16: abstract 497, 2002 (available on CD-ROM).

40

Caspers S, Geyer S, Schleicher A, Mohlberg H, Amunts K, Zilles K. The human

inferior parietal cortex: cytoarchitectonic parcellation and interindividual

variability. Neuroimage 33: 430-448, 2006.

Chao LL, Martin A. Representation of manipulable man-made objects in the

dorsal stream. Neuroimage 12: 478-484, 2000.

Cohen MS, Kosslyn SM, Breiter HC, DiGirolamo GJ, Thompson WL, Anderson

AK, Brookheimer SY, Rosen BR, Belliveau JW. Changes in cortical

activity during mental rotation. A mapping study using functional MRI.

Brain 119 ( Pt 1): 89-100, 1996.

Creem-Regehr SH, Lee JN. Neural representations of graspable objects: are

tools special? Brain Res Cogn Brain Res 22: 457-469, 2005.

Culham J C, Danckert SL, DeSouza JF, Gati JS, Menon RS, Goodale MA.

Visually guided grasping produces fMRI activation in dorsal but not

ventral stream brain areas. Exp Brain Res 153: 180-189, 2003.

Culham JC, Cavina-Pratesi C, Singhal A. The role of parietal cortex in

visuomotor control: what have we learned from neuroimaging?

Neuropsychologia 44: 2668-2684, 2006.

41

Decety J, Perani D, Jeannerod M, Bettinardi V, Tadary B, Woods R, Mazziotta

JC, Fazio F. Mapping motor representations with positron emission

tomography. Nature 371: 600-602, 1994.

Duvernoy HM. The human brain. Surface, blood supply, and three dimensional

sectional anatomy. New York: Springer, 1999.

Fadiga L, Fogassi L, Gallese V, Rizzolatti G. Visuomotor neurons: ambiguity of

the discharge or 'motor' perception? Int J Psychophysiol 35: 165-177,

2000.

Fagg AH, Arbib MA. Modeling parietal-premotor interactions in primate control

of grasping. Neural Netw 11: 1277-1303, 1998.

Fogassi L, Ferrari PF, Gesierich B, Rozzi S, Chersi F, Rizzolatti G. Parietal

lobe: from action organization to intention understanding. Science 308:

662-667, 2005.

Frey SH, Vinton D, Norlund R, Grafton ST. Cortical topography of human

anterior intraparietal cortex active during visually guided grasping.

Brain Res Cogn Brain Res 23: 397-405, 2005.

Friston KJ, Holmes AP, Worsley KJ, Poline JB, Frith CD, Frackowiak RSJ.

42

Statistical parametric maps in functional imaging: a general linear

approach. Hum Brain Mapp 2: 189-210, 1995.

Friston KJ, Buechel C, Fink GR, Morris J, Rolls E, Dolan RJ.

Psychophysiological and modulatory interactions in neuroimaging.

NeuroImage 6: 218-229, 1997.

Friston KJ, Penny WD, Glaser DE. Conjunction revisited. Neuroimage 25:

661-667, 2005.

Garraux G, McKinney C, Wu T, Kansaku K, Nolte G, Hallett M. Shared brain

areas but not functional connections controlling movement timing and

order. J Neurosci 25: 5290-5297, 2005.

Genovese CR, Lazar NA, Nichols T. Thresholding of statistical maps in

functional neuroimaging using the false discovery rate. Neuroimage 15:

870-878, 2002.

Gitelman DR, Penny WD, Ashburner J, Friston KJ. 2003. Modeling regional

and psychophysiologic interactions in fMRI: the importance of

hemodynamic deconvolution. Neuroimage 19: 200-207, 2003.

Grafton ST, Arbib MA, Fadiga L, Rizzolatti G. Localization of grasp

43

representations in humans by positron emission tomography. 2.

Observation compared with imagination. Exp Brain Res 112: 103-111,

1996a.

Grafton ST, Fadiga L, Arbib MA, Rizzolatti G. Premotor cortex activation

during observation and naming of familiar tools. Neuroimage 6: 231-236,

1997.

Grafton ST, Fagg AH, Woods RP, Arbib MA. Functional anatomy of pointing and

grasping in humans. Cereb Cortex 6: 226-237, 1996b.

Grefkes C, Fink GR. The functional organization of the intraparietal sulcus in

humans and monkeys. J Anat 207: 3-17, 2005.

Greicius MD, Krasnow B, Reiss AL, Menon V. Functional connectivity in the

resting brain: a network analysis of the default mode hypothesis. Proc

Natl Acad Sci U S A 100: 253-258, 2003.

Grezes J, Armony JL, Rowe J, Passingham RE. Activations related to "mirror"

and "canonical" neurones in the human brain: an fMRI study.

Neuroimage 18: 928-937, 2003.

Grill-Spector K. The neural basis of object perception. Curr Opin Neurobiol

44

13:159-166, 2003.

Gusnard DA, Raichle ME. Searching for a baseline: functional imaging and the

resting human brain. Nat Rev Neurosci 2: 685-694, 2001.

Heilman KM, Rothi LJ, Valenstein E. Two forms of ideomotor apraxia.

Neurology 32: 342-346, 1982.

Jeannerod M. Motor Cognition: What Actions Tell the Self. Oxford, U.K: Oxford

University Press, 2006, p. 12-16.

Jeannerod M, Arbib MA, Rizzolatti G, Sakata H. Grasping objects: the cortical

mechanisms of visuomotor transformation. Trends Neurosci 18: 314-320,

1995.

Jeannerod M, Decety J, Michel F. Impairment of grasping movements following

a bilateral posterior parietal lesion. Neuropsychologia 32 : 369-380, 1994.

Johnson SH, Rotte M, Grafton ST, Hinrichs H, Gazzaniga MS, Heinze HJ.

Selective activation of a parietofrontal circuit during implicitly imagined

prehension. Neuroimage 17: 1693-1704, 2002.

Kosslyn SM, DiGirolamo GJ, Thompson WL, Alpert NM. Mental rotation of

objects versus hands: neural mechanisms revealed by positron emission

45

tomography. Psychophysiology 35: 151-161, 1998.

Knudsen EI. Fundamental Components of Attention. Annual Review of

Neuroscience. 30: 57-78, 2007.

Mazoyer B, Zago L, Mellet E, Bricogne S, Etard O, Houde O, Crivello F, Joliot M,

Petit L, Tzourio-Mazoyer N. Cortical networks for working memory and

executive functions sustain the conscious resting state in man. Brain Res

Bull 54: 287-298, 2001.

Murata A, Fadiga L, Fogassi L, Gallese V, Raos V, Rizzolatti G. Object

representation in the ventral premotor cortex (area F5) of the monkey. J

Neurophysiol 78: 2226-2230, 1997.

Nichols T, Brett M, Andersson J, Wager T, Poline JB. 2005. Valid conjunction

inference with the minimum statistic. Neuroimage 25: 653-660, 2005.

Ogawa S, Menon RS, Tank DW, Kim SG, Merkle H, Ellermann JM, Ugurbil K.

Functional brain mapping by blood oxygenation level-dependent contrast

magnetic resonance imaging. A comparison of signal characteristics with

a biophysical model. Biophys J 64: 803-812, 1993.

Oldfield RC. The assessment and analysis of handedness: the Edinburgh

46

inventory. Neuropsychologia 9: 97-113, 1971.

Ono M, Kubik S, Abernathey CD. Atlas of the Cerebral Sulci. New York: Thieme

Medical Publishers, Inc, 1990.

Pandya DN, Seltzer B. Intrinsic connections and architectonics of posterior

parietal cortex in the rhesus monkey. J Comp Neurol 204: 196-210, 1982.

Penny WD, Holmes AP. Random effect analysis. In: Human Brain Function.

Second edition, edited by Frackowiak RSJ, Friston KJ, Frith CD, Dolan

RJ, Price CJ, Zeki S, Ashburner J, Penny W, New York: Academic Press,

2003, p. 843-850.

Petrides M, Pandya DN. Projections to the frontal cortex from the posterior

parietal region in the rhesus monkey. J Comp Neurol 228: 105-116, 1984.

Richter W, Somorjai R, Summers R, Jarmasz M, Menon RS, Gati JS,

Georgopoulos AP, Tegeler C, Ugurbil K, Kim SG. Motor area activity

during mental rotation studied by time-resolved single-trial fMRI. J Cogn

Neurosci 12: 310-320, 2000.

Rizzolatti G, Camarda R, Fogassi L, Gentilucci M, Luppino G, Matelli M.

Functional organization of inferior area 6 in the macaque monkey. II.

47

Area F5 and the control of distal movements. Exp Brain Res 71: 491-507,

1998.

Rizzolatti G, Luppino G. The cortical motor system. Neuron 31: 889-901, 2001.

Rizzolatti G, Matelli M. Two different streams form the dorsal visual system:

anatomy and functions. Exp Brain Res 153: 146-157,2003.

Rozzi S, Calzavara R, Belmalih A, Borra E, Gregoriou GG, Matelli M, Luppino

G. Cortical connections of the inferior parietal cortical convexity of the

macaque monkey. Cereb Cortex 16: 1389-1417, 2006.

Rozzi S, Ferrari PF, Bonini L, Rizzolatti G, Fogassi L. Functional organization

of inferior parietal lobule convexity in the macaque monkey:

electrophysiological characterization of motor, sensory and mirror

responses and their correlation with cytoarchitectonic areas. Eur J

Neurosci 28: 1569-1588, 2008.

Sakata H. 2003. The role of the parietal cortex in grasping. Adv Neurol 93:

121-139, 2003.

Shikata E, Hamzei F, Glauche V, Koch M, Weiller C, Binkofski F, Buchel C.

Functional properties and interaction of the anterior and posterior

48

intraparietal areas in humans. Eur J Neurosci 17: 1105-1110, 2003.

Shulman GL, Fiez JA, Corbetta M, Buckner RL, Miezin FM, Raichle ME,

Petersen SE. Common Blood Flow Changes across Visual Tasks: II.

Decreases in Cerebral Cortex. J Cogn Neurosci 9: 648-663, 1997.

Snodgrass JG, Vanderwart M. A standardized set of 260 pictures: norms for

name agreement, image agreement, familiarity, and visual complexity. J

Exp Psychol [Hum Learn] 6: 174-215, 1980.

Taira M, Mine S, Georgopoulos AP, Murata A, Sakata H. Parietal cortex

neurons of the monkey related to the visual guidance of hand movement.

Exp Brain Res 83: 29-36, 1990.

Talairach J, Tournoux P. Coplanar stereotaxic atlas of the human brain.

3-Dimensional proportional system: an approach to cerebral imaging.

New York: Thieme Medical Publishers, Inc, 1988.

Tunik E, Rice NJ, Hamilton A, Grafton ST. Beyond grasping: representation of

action in human anterior intraparietal sulcus. Neuroimage 36 Suppl 2:

T77-86, 2007.

Ungerleider LG, Mishkin M. Two cortical visual systems. In: Analysis of Visual

49

Behavior, edited by Ingle DJ, Goodale MA, Mansfield RJW. Cambridge,

MA: The MIT Press, 1982.

Valyear KF, Cavina-Pratesi C, Stiglick AJ, Culham JC. Does tool-related fMRI

activity within the intraparietal sulcus reflect the plan to grasp?

Neuroimage 36 S2: T94-T108, 2007.

Wagner AD, Schacter DL, Rotte M, Koutstaal W, Maril A, Dale AM, Rosen BR,

Buckner RL. Building memories: remembering and forgetting of verbal

experiences as predicted by brain activity. Science 281: 1188-1191, 1998.

Wheaton LA, Hallett M. Ideomotor apraxia: a review. J Neurol Sci 260: 1-10,

2007.

50

Table1

Line-drawn pictures adopted for the present fMRI experiment and the results of

the subjects’ judgment for each object as to whether graspable or not.

Object ID number* Order of

presentation**

Ratio of the judgment as

"graspable"***

Functional

categories*

***

Apple

6 2 1 G-F

Asparagus

11 4 1 G-F

Banana

16 9 1 G-F

Carrot

48 24 1 G-F

Celery

51 25 1 G-F

51

Cherry

54 29 1 G-F

Cigar

58 32 1 G-F

Cigarette

59 34 1 G-F

Corn

66 39 1 G-F

Axe

12 5 1 G-SO

Baseball bat

19 12 1 G-SO

Basket

20 14 1 G-SO

Bowl

34 15 1 G-SO

Brush

38 19 1 G-SO

Button

41 22 1 G-SO

52

Clothespin

61 35 1 G-SO

Comb

65 38 1 G-SO

Cup

70 42 1 G-SO

Doorknob

77 45 1 G-SO

Football

95 48 1 G-SO

Fork

97 49 1 G-SO

Bread

36 18 0.94 G-F

Ball

14 8 0.94 G-SO

Chain

52 28 0.94 G-SO

Doll

74 44 0.65 Amb

53

Chair

53 21 0.53 Amb

Cat

49 20 0.47 Amb

Accordion

1 1 0.35 Amb

Chicken

55 23 0.29 Amb

Eye

86 41 0.29 Amb

Dog

73 36 0.12 NG-A

Fox

98 46 0.12 NG-A

Alligator

3 3 0.06 NG-A

Bear

21 7 0.06 NG-A

Bed

22 10 0.06 NG-LO

54

Cannon

45 16 0.06 NG-LO

Car

47 17 0.06 NG-LO

Church

57 26 0.06 NG-LO

Fence

87 43 0.06 NG-LO

Camel

43 13 0 NG-A

Cow

68 30 0 NG-A

Deer

71 31 0 NG-A

Donkey

75 37 0 NG-A

Elephant

84 40 0 NG-A

Giraffe

103 47 0 NG-A

55

Goat

107 50 0 NG-A

Barn

17 6 0 NG-LO

Bus

39 11 0 NG-LO

Couch

67 27 0 NG-LO

Desk

72 33 0 NG-LO

* ID number in the list of Snodgrass and Vanderwart (1980).

** Order of presentation for each object during the fMRI experiment.

*** Ratio of the subjects who judged “graspable” among the total number of

subjects (n=17). For each object, if all subjects judged “graspable”, the ratio is

1.00, and if none of the subjects judged “graspable”, the ratio is 0.00.

**** Functional categories: G-F: food or objects graspable by one hand or placed

in mouth, G-SO: graspable small manufactured or round objects, Amb: objects

with ambiguous graspability, NG-A: non-graspable animals, NG-LO:

56

non-graspable large and/or immobile objects. See Results section for the detail

of categorization.

57

Table 2

Brain areas commonly activated in the judgment of graspable and

non-graspable objects, shown as the results of conjunction analysis.

Activated areas Brodmann area x y z T Z score

Left cortical areas

Superior frontal sulcus,

PMd

6 -26 -4 54 6.77 5.29

Precentral gyrus, M1 4 -54 -8 50 4.25 3.76

Precentral gyrus, PMd 6 -48 -4 56 4.93 4.22

Precentral gyrus, PMv 6 -46 2 38 5.23 4.42

Middle frontal gyrus, DLPF 46 -48 30 22 5.77 4.74

9 -40 8 28 4.76 4.11

10 -30 46 22 4.46 3.91

8 -52 10 44 4.08 3.63

Inferior frontal gyrus, VLPF 45 -32 24 -4 6.40 5.10

58

44 -54 6 24 4.66 4.04

Inferior frontal gyrus, OF 10 -50 42 0 4.42 3.87

Insula 13 -34 22 4 6.30 5.04

Superior parietal lobule 7 -28 -64 48 6.87 5.35

Inferior parietal lobule 40 -42 -36 42 4.71 4.08

Superior temporal gyrus 22 -50 -48 8 2.78 2.61

Middle temporal gyrus 21 -64 -50 -4 3.33 3.06

Inferior temporal gyrus 20 -56 -56 -16 5.31 4.46

Parahippocampal gyrus 36 -36 -24 -28 4.56 3.98

27 -22 -34 -8 3.14 2.91

Middle occipital gyrus 18 -38 -82 -14 7.11 5.47

19 -28 -80 28 5.39 4.52

Inferior occipital gyrus 18 -24 -90 -14 5.04 4.29

Cuneus 17 -12 -86 4 3.07 2.85

Lingual gyrus 17 -12 -102 -10 4.94 4.23

18 -12 -74 2 3.10 2.87

59

Fusiform gyrus 37 -40 -44 -26 9.30 6.43

19 -44 -70 -16 7.59 5.70

Right cortical areas

Precentral gyrus, PMd 6 50 -6 58 3.23 2.98

Precentral gyrus, PMv 6 50 0 40 5.39 4.52

Middle frontal gyrus, PMv 6 50 6 46 5.28 4.45

Middle frontal gyrus, DLPF 8 42 6 48 4.88 4.19

8/9 48 10 34 4.77 4.12

10 40 46 16 4.20 3.72

46 40 32 28 3.91 3.51

Inferior frontal gyrus, VLPF 45 38 22 0 7.57 5.69

44/45 60 26 18 4.63 4.02

44 44 14 18 4.54 3.96

Inferior frontal gyrus, DLPF 10 48 50 0 3.71 3.36

Inferior frontal gyrus, OF 11 26 36 -20 3.51 3.20

60

Insula 13 38 16 10 5.12 4.34

Superior parietal lobule 7 34 -68 42 6.20 4.99

Inferior parietal lobule 40 54 -36 48 4.09 3.64

Superior temporal gyrus 22 66 -44 10 2.77 2.60

Middle temporal gyrus 37 58 -46 -10 4.51 3.94

21 54 -46 8 3.31 3.04

Hippocampus 28 -26 -8 3.31 3.04

Parahippocampal gyrus 19 30 -46 -10 3.80 3.43

Middle occipital gyrus 19 34 -90 14 5.10 4.33

18 32 -84 -16 4.96 4.24

Cuneus 17 14 -82 6 3.82 3.44

Lingual gyrus 17 16 -102 -6 6.46 5.13

18 26 -74 -2 4.03 3.60

Fusiform gyrus 20 30 -40 -26 10.50 6.86

37 44 -52 -22 7.57 5.69

19 42 -66 -16 7.50 5.66

61

Midline cortical areas

Left preSMA 6 0 6 60 8.46 6.09

Right preSMA 6 8 10 52 8.75 6.21

Left SMA 6 -4 -2 66 8.26 6.00

Left anterior cingulate

gyrus

24 -6 26 36 5.32 4.47

Right anterior cingulate

gyrus

24 12 16 38 6.96 5.39

Left posterior cingulate

gyrus

23 -4 -34 24 3.89 3.50

Subcortical areas

Left lentiform nucleus -18 -2 0 3.53 3.22

Right lentiform nucleus 16 2 0 3.39 3.11

Left thalamus -14 -14 4 3.57 3.25

62

Right thalamus 12 -12 2 3.28 3.02

Left cerebellum -36 -54 -32 7.00 5.41

Right cerebellum 38 -46 -34 6.60 5.20

PMd, dorsal premotor cortex; PMv, ventral premotor cortex; M1, primary motor

cortex; DLPF, dorsolateral prefrontal cortex; VLPF, ventrolateral prefrontal

cortex; OF, orbiofrontal cortex; SMA, supplementary motor area.

Brodmann areas and coordinates (in mm) of peaks in MNI space are given for

information. Anatomic localization was performed with the aid of the atlas by

Duvernoy (1999).

63

Table 3

Brain areas more activated in the judgment of graspable objects as compared

with that of non-graspable objects.

Cluster

Cluster

size

Activated areas

Brodmann

area

x y z T Z score

1 496

Left ventral inferior

parietal lobule

40 -62 -32 40 7.33 4.79

2 231

Left dorsal inferior

parietal lobule

40 -46 -44 60 5.53 4.08

Cluster size is the number of voxels. Brodmann areas and coordinates (in mm)

of peaks of clusters in MNI space are given for information. Cluster number

corresponds to that in Figure 2.

64

Table 4

Brain areas showing differential activation among different functional

categories.

Functional

categories*

Cluster

size Activated areas

Brodmann

area x y z T

Z

score

NG-LO vs.NG-A 213 Left fusiform

gyrus 37 -22 -46 -10 6.76 4.58

150 Right fusiform

gyrus 37 28 -38 -18 6.19 4.36

G-SO vs. NG-LO 112 Left inferior

temporal gyrus 37 -54 -52 -24 5.38 4.01

NG-LO vs. G-F 203 Left fusiform

gyrus 19 -28 -48 -10 5.52 4.07

115 Right fusiform

gyrus 37 30 -38 -20 6.97 4.66

107 Left middle

occipital gyrus 19 -34 -88 14 6.2 4.36

G-F vs. NG-LO 125 Left inferior

parietal lobule 40 -62 -36 42 4.9 3.77

*Functional categories: NG-LO: non-graspable large and/or immobile objects,

NG-A: non-graspable animals, G-SO: graspable small manufactured or round

objects, G-F: food or objects graspable by one hand or placed in mouth.

65

Table 5

Brain areas showing significant psychophysiological interaction (PPI) from the

ROIs in the IPL when subjects judge presented objects as graspable compared

with non-graspable.

Index areas

Activated

areas

x y z T Z score

Left ventral inferior parietal lobule Left DLPF -46 10 42 4.95 3.80

Left/right

pre-SMA

10 10 56 4.84 3.74

Left dorsal inferior parietal lobule

Right

cerebellum

22 -42 -18 5.84 4.22

Left PMv -58 2 26 5.30 3.97

Left/right

SMA

-10 0 68 4.18 3.39

DLPF, dorsolateral prefrontal cortex; SMA, supplementary motor area; PMv,

66

ventral premotor cortex. Coordinates (in mm) of peaks in MNI space are given

for information. Anatomic localization was performed with the aid of the atlas

by Duvernoy (1999).

67

Figure legends

Figure 1

Brain areas showing common activation in the judgment of graspable and

non-graspable objects. Results were thresholded at p < 0.05 corrected for

multiple comparisons and rendered on a standard brain image. Brain areas

significantly activated compared with the resting condition include bilateral

dorsal and ventral premotor cortex, bilateral dorsolateral prefrontal cortex,

inferior frontal gyrus, pre-supplementary motor area, superior and inferior

parietal lobules, and fusiform gyrus.

Figure 2

Brain areas more activated in the judgment of graspable objects than that of

non-graspable objects. Results were thresholded at p < 0.001 (uncorrected) and

rendered on a standard brain image for demonstration (top). Two clusters both

located in the left inferior parietal lobule (IPL) were activated significantly at p

< 0.05 (corrected) (large arrows). A cluster in the premotor cortex, the peak of

68

which was chosen as a seed for the subsequent region of interest analysis, is

shown as well (small arrow). The images in the middle are a series of coronal

sections of a standard brain to show the activation in the left IPL. The

consecutive activation below the white dashed line compose the cluster located

ventrally in the IPL (ventral IPL) and the consecutive activation above the line

compose the cluster located dorsally in the IPL (dorsal IPL). The color bar

represents the T score. On the bottom panel, parameter estimates (90%

confidence interval) of the peaks of the two clusters showing activation

associated with judgment of graspable objects (G-judgment) and non-graspable

objects (N-judgment) are shown. The peak of the ventral IPL cluster (large

arrow 1) was activated significantly in both the G-judgment and N-judgment

conditions compared with the resting state (bottom left). By contrast, the peak

of the dorsal IPL cluster (large arrow 2) showed significant activation only in

the G-judgment condition (bottom right) (p < 0.05, corrected).

Figure 3

Differential signal changes in regions of interest (ROIs) in the left dorsal

69

inferior parietal lobule (IPL), ventral IPL and ventral premotor cortex (PMv).

Percent signal changes associated with judgment of non-graspable objects

(N-judgment) were subtracted from those associated with judgment of

graspable objects (G-judgment). Repeated measures ANOVA shows significant

difference in the signal changes among these three ROIs (F (1.5/23.996)=10.093,

p < 0.005). Post-hoc pairwise comparison test demonstrates that signal

changes in the dorsal IPL ROI are significantly larger than those in the ventral

IPL and PMv ROIs (p < 0.005). The error bars characterize standard

deviations.

Figure 4

Brain areas showing significant psychophysiological interaction (PPI) from the

regions of interest (ROIs) in the inferior parietal lobule (IPL) shown in Figure 2

when subjects judged presented objects as graspable compared with

non-graspable. In the PPI analysis with the ventral IPL ROI (A), the left

dorsolateral prefrontal cortex and pre-supplementary motor area show

significant PPI, and in the PPI analysis with the dorsal IPL ROI (B), the left

70

ventral premotor cortex, supplementary motor area and right cerebellum have

significant PPI.

Figure 5

Fused rendering images of the results of PPI analysis (Figure 4). The brain

areas which have significant positive modulation in the functional connectivity

from the regions of interest (ROI) in the inferior parietal lobule (IPL) when

subjects judged presented objects as graspable compared with non-graspable

are rendered on a standard brain image. Red, target areas from the ventral

IPL (Figure 4A); green, target areas from the dorsal IPL (Figure 4B). These

two IPL ROIs have similar functional connectivity over the frontal cortex, but,

as these fused images show, the connectivity from the two areas is clearly

discrete.

G-judgment

0

0.4

0.8

1.2

1.6

2.0

2.4

-0.4

0.4

0.8

1.2

1.6

2.0

2.4

0

2.8

Greater Activation in G- over N- Judgment

Ventral IPL [-62, -32, 40] Dorsal IPL [-46, -44, 60]

N-judgment

G-judgment

N-judgment

1

2

Y = -24 -28 -32 -36 -40 -44 -48 -52

PPI with the left ventral IPL

L R

PPI analysis red: left ventral IPL green: left dorsal IPL