Skip to main content
NCBI home page
Human Brain Mapping logoLink to Human Brain Mapping
. 2011 Jul 7;33(8):1821–1833. doi: 10.1002/hbm.21323

Hand shape selection in pantomimed grasping: Interaction between the dorsal and the ventral visual streams and convergence on the ventral premotor area†

Michiru Makuuchi 1,2,✉, Yoshiaki Someya 3, Seiji Ogawa 3, Yoshihiro Takayama 2
PMCID: PMC3193865  NIHMSID: NIHMS290745  PMID: 21739528

Abstract

In visually guided grasping, possible hand shapes are computed from the geometrical features of the object, while prior knowledge about the object and the goal of the action influence both the computation and the selection of the hand shape. We investigated the system dynamics of the human brain for the pantomiming of grasping with two aspects accentuated. One is object recognition, with the use of objects for daily use. The subjects mimed grasping movements appropriate for an object presented in a photograph either by precision or power grip. The other is the selection of grip hand shape. We manipulated the selection demands for the grip hand shape by having the subjects use the same or different grip type in the second presentation of the identical object. Effective connectivity analysis revealed that the increased selection demands enhance the interaction between the anterior intraparietal sulcus (AIP) and posterior inferior temporal gyrus (pITG), and drive the converging causal influences from the AIP, pITG, and dorsolateral prefrontal cortex to the ventral premotor area (PMv). These results suggest that the dorsal and ventral visual areas interact in the pantomiming of grasping, while the PMv integrates the neural information of different regions to select the hand posture. The present study proposes system dynamics in visually guided movement toward meaningful objects, but further research is needed to examine if the same dynamics is found also in real grasping. Hum Brain Mapp, 2012. © 2011 Wiley Periodicals, Inc

Keywords: functional magnetic resonance imaging, grip selection, object recognition, effective connectivity

INTRODUCTION

Neural mechanisms of visuomotor transformation in grasping have been uncovered in detail [Binkofski et al.,1998; Culham et al.,2003; Ehrsson et al.,2000; Fogassi et al.,2001; Gallese et al.,1994; Grol et al.,2007; Luppino et al.,1999; Murata et al.,1997,2000; Raos et al.,2006; Rizzolatti et al.,1988; Sakata et al.,1995; Verhagen et al.,2008], but we still poorly understand how we select one grasping hand shape from the multiple possibilities by integrating the object meaning and the goal of the action [Castiello and Begliomini,2008; Cisek and Kalaska,2010; Fagg and Arbib,1998; Rizzolatti and Luppino,2001]. The present fMRI study aimed to reveal the system dynamics of the brain during the pantomiming of grasping hand shape formation in humans using dynamic causal modeling (DCM) analysis [Friston et al.,2003].

The subjects pantomimed grasping an object either by pinching with the thumb and index finger (precision grip) or by gripping it between the palm and all the fingers and thumb (power grip) according to symbols drawn on photographs of graspable objects used in everyday life (see Fig 1). The experimental tasks were designed to emphasize two aspects in grasping. One is the interaction of the visuomotor transformation in the dorsal visual stream and the object recognition in the ventral visual stream [Goodale and Milner,1992; Ungerleider and Haxby,1994] evoked by the use of objects for daily use. Given that grasping an object is often the first action in goal‐directed behavior, semantic memory retrieval of the target object (such as the purpose, use, and various intrinsic properties of the object) can be involved in the hand preshaping. Indeed, a recent anatomical study in macaques reported the connection between the anterior intra parietal area (AIP) and the ventral visual areas [Borra et al.,2008], suggesting their interaction is highly likely.

Figure 1.

Figure 1

Open in a new tab

Examples of the stimuli and the grips. The diagram at the bottom represents task blocks. In each block, seven objects are shown with duration of 4 s. Here we show examples of the stimuli (middle row) and the responses (top row). In the PRE1 condition, an iron kettle and a wrench are presented. One or two asterisks are drawn on the pictures to indicate where to put the index finger and the thumb in a precision grip. In PRE2same condition, the same picture (the wrench with two asterisks) is shown again and the subject performs the same miming of grasping. By contrast, the same objects (an iron kettle) is shown in POW2diff condition but with different symbols (a bar), which indicates the part to be grasped with the power grip. The subject is requested to change the grip type in “diff” conditions as opposed to repeating the same grip in “same” conditions. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

The other aspect of emphasis is the selection of the grip hand shape. To enhance the neural activities for selection, we employed a psychological technique that increases or decreases the selection demand in the hand grip formation. Crucially, each object was shown twice in the same session and the grip in the second presentation of the objects was the same as or different from the initial grip type (i.e., precision or power grip). When the same grip is requested for the identical object, brain activation for the second movement is expected to be smaller than for the first, as in the neural adaptation observed in various modalities [Chong et al.,2008; Grill‐Spector et al.,2006; Hamilton and Grafton,2006; Kilner et al.,2009; Koutstaal et al.,2001; Shmuelof and Zohary,2005]. By contrast, when a different grip is requested in the second presentation, the implicit memory of the first response to this object will compete with the current grip selection, thereby increasing the activity in areas relevant to grip selection, as shown in previous studies in the language domain [Thompson‐Schill et al.,1997,1999]. We expected that the increased selection demands would enhance both the regional activities and the inter‐regional dynamic causal relations in the selection of the grip.

METHOD

Subjects

Sixteen young, right‐handed, healthy subjects were examined (nine males and six females). Handedness was assessed with the Edinburgh Inventory [males: mean 97.9, range 79–100, females: mean 100, Oldfield, 1971]. For males, the mean age was 22.4‐years old and the range was 19‐ to 27‐years old, and the mean 22.3 and the range 19–24 for females. All had no history of neurological disorders. The experimental procedures were approved by the Research Ethics Committees of the University of Tokyo and by that of Hamano Life Science Research Foundation Ogawa Laboratories for Brain Function Research. Written informed consent was given by all subjects.

Stimuli

Subjects were presented color photographs of 168 common objects used or found at home, such as tools, cooking utensils, pans, cups, saucers, toys, vegetables, fruits, food, bottles, tubes, cans, stationary, bags, accessories etc (see Fig. 1). The photographs were taken against a black background. Trademarks or logos on the objects were blurred to make them unreadable. All the objects were projected onto a screen in the size of 5 × 5° so that information about the size of the objects was not available in the images. The subjects were informed that all the objects were a graspable size, but they needed to recognize the objects in order to make the correct grasping movements, especially for the apertures between the thumb and the index finger (precision grip) or between the thumb and other fingers (power grip).

EXPERIMENTAL CONDITIONS

The First Precision Grip Condition (PRE1)

The subjects were instructed to mime the precision grip movement with the right hand as if the object projected onto the screen was close to their right hand so that they did not need to reach for it. Subjects were instructed to make the aperture between the index finger and the thumb, the grip force, and the orientation of the wrist as appropriate as possible for the objects presented. In order to help them to carry out the grasping movements without the objects, we advised them to recall/imagine the object properties such as size, shape, weight, texture of the surface, purposes of use, how to use etc. The position to place the index finger and the thumb was indicated by one or two yellow asterisks (*) drawn on the picture (see Fig. 1). All the objects presented in this condition were new to them.

The First Power Grip Condition (POW1)

The subjects were instructed to mime the grasping movement to the presented objects in the picture by opposing the thumb and all fingers to the palm of the right hand without a reaching movement. Except for the type of grip, the subjects were instructed to perform the task the same as in the PRE1 condition. The position to put the palm was indicated by a yellow bar symbol (see Fig. 1) drawn on the picture. Since no indication as to where to put the thumb and fingers was provided, the subjects had to decide this by themselves. All the objects presented in this condition were new to them.

The Second Precision Grip Condition (PRE2same)

The subjects performed the same precision grip movements as in the PRE1 condition. The difference between the PRE1 and the PRE2same conditions was that the objects presented in the PRE2same had already appeared in the PRE1 so that the subjects made the same grasping movements to the objects seen and gripped previously.

The Second Power Grip Condition (POW2same)

Equivalent to the PRE2same condition, the subjects performed the power grip movements again with the objects already presented in the POW1 condition.

The Precision Grip With Preceding Power Grip Condition (PRE2diff)

The subjects performed the precision grip movements as in the PRE1 condition, but the objects to be grasped had already been presented in the preceding POW1 condition. Therefore, the subjects made precision grip hand movements to the objects for the first time, but they had already made power grip movements to the same objects in the preceding blocks.

The Power Grip With Preceding Precision Grip Condition (POW2diff)

This condition was the opposite of the PRE2diff condition. The subjects performed a power grip with the objects that had been presented and processed in the PRE1 condition.

Baseline Condition

The subjects were instructed to fixate their eyes on a small white point located in the center of a black background.

Experimental Design

The experiments were designed as block‐design fMRI experiments. The subjects had three separate sessions with 56 distinct different objects assigned to each session. In a session, half of the 56 objects were used in the PRE1 condition and the rest were in the POW1 condition. Half of these 28 objects used in the PRE1 condition were presented in the PRE2same condition, and the other half were in the POW2diff condition. Similarly, the 28 objects appearing in the POW1 condition were divided into two halves, one for the POW2same and the other for the PRE2diff. One trial had a duration of 4 s and the picture of the objects were presented for 4 s. Seven trials, with a 2 s blank screen after the last trial, comprised one block, with a total duration of 30 s. One session had a total of 20 blocks, which consisted of four PRE1, four POW1, two PRE2same, two PRE2diff, two POW2diff, two POW2same and four baseline conditions. The baseline condition was presented every fifth block. The order of the conditions was randomized between three sessions, but blocks for the PRE1 and the POW1 conditions tended to concentrate in the early halves of the sessions due to the definition of the conditions.

Experimental Procedures

The subjects were instructed on the tasks before the fMRI sessions. The precision grip and power grip were demonstrated by the experimenter to them using a few objects that did not appear in the fMRI experiments. None of the objects shown in the fMRI sessions were presented to the subjects before scanning. Stimulus presentation was programmed with Presentation 0.80 software (Neurobehavioral Systems, Albany, CA) on a Windows notebook PC (Dynabook G5/X16PME, Toshiba, Tokyo, Japan). Stimuli were projected through an LCD projector (PRO Xtrax LP‐XF31 Multiverse projector, SANYO, Tokyo, Japan) onto the back of a screen. The subjects viewed the images on the screen that stood above the head through a mirror attached to the head‐coil. To prevent head motion, a vacuum pillow or a viscoelastic foam pillow was laid under the head. Hand movements were recorded by video from the outside of the scanner room through the window for off‐line evaluation of performance. To make the hand visible to the video camera, foam pads were put under the right forearm while keeping the elbow in contact with the bed.

Image Acquisition

EPI data were acquired with a Siemens Allegra operating at 3.0 T (Siemens Medical Solutions, Erlangen, Germany). The whole brain was covered with 5.0‐mm thick 20 axial images with 1.0‐mm gaps (TR = 1.2 s, TE = 30 ms, Flip angle = 80°, FOV = 24 × 24 cm2, 64 × 64 matrix). The subjects had three sessions of fMRI scanning with 480 volumes per session. A total of 1440 volumes were collected in about 40 min. Structural high‐resolution images of the subjects' heads were also collected between the experimental sessions with a three‐dimensional MPRAGE sequence (TR = 2800 ms, TE = 4.38 ms, TI = 1,300 ms, Flip angle = 8°, FOV = 24 × 24 cm2, 256 × 256 matrix, axial 128 slices, 1.3‐mm thick).

Preprocessing of Structural and Functional MRI Data

The first five volumes of each fMRI session were discarded to eliminate magnetic saturation effects, and a total of 475 volumes per session were used. The data analysis was carried out using SPM8 (available at http://www.fil.ion.ucl.ac.uk/spm/) on Linux PC workstations. Structural images were normalized using the DARTEL procedure [Ashburner,2007], in which individual structural images are segmented into gray and white matter, and a mean image from each individual's image serves as templates. The DARTEL normalization proceeds in six steps with increasing spatial resolution, with the final step for the linear transformation into MNI space. This process yielded normalized individual structural images in 1 × 1 × 1 mm3 resolution. For the later use of individual normalized images as the target images, the images were smoothed with a 2‐mm FWHM Gaussian kernel. For preprocessing of the functional data, EPI images were realigned to the first image and resliced. Then the difference in the slice acquisition time was corrected and the volumes were resliced again. We normalized individual structural and functional data using the individual structural images that were normalized by DARTEL as the target images. The functional images were normalized with resampling at 3 × 3 × 3 mm3 and smoothed with a 6‐mm FWHM Gaussian kernel.

Within‐Subject ANOVA of Functional MRI Data

Each subject's hemodynamic responses induced by the task blocks were modeled with a box‐car function with duration of 30 s and convolved with a hemodynamic function that reaches a peak 6.0 s after the stimuli onset. The global mean intensity of each session was normalized to 100. Confounds by global signal changes were removed by applying a high pass filter with a cut‐off cycle of 128 s. Signal increase relative to the baseline in the POW1 and the PRE1 of each participant was estimated according to the general linear model. The signal differences between the second and the first presentation were estimated by the contrast of POW2same‐POW1, POW2diff‐ POW1, PRE2same‐ PRE1, and PRE2diff‐ PRE1. These four difference contrast images of all subjects were submitted to the second level (group) analysis, a 2 × 2 within subject ANOVA with factors GRIP‐CHANGE (DIFFERENT/SAME) and TYPE (POW/PRE) with correction for nonsphericity. Main effects and interactions were tested with t tests since we were interested in the effects with the direction. Statistical inferences were drawn at P < 0.05 at cluster level: The statistical maps (SPM{T}) were thresholded at P < 0.001 (not corrected) for intensity, and then thresholded by cluster size (50 contiguous voxels). We also calculated the “effect of interest” contrast that is an F tests on the four difference contrast images with the null hypothesis that all the images are equal to zero. This test reveals voxels in which at least one of the difference contrasts images are significantly non‐zero.

Dynamic Causal Modeling

VOI selection

To understand the effective connectivity between the activated regions, we performed a dynamic causal model (DCM) analysis [Friston et al.,2003]. We constructed DCM models with the five volumes of interest (VOIs), namely the ventral premotor area (PMv), anterior intraparietal area (AIP), dorsolateral prefrontal cortex (DLPFC), posterior inferior temporal gyrus (pITG), and primary visual area (V1) in the left hemisphere were selected to construct 28 alternative DCMs. The rationale for selecting these five VOIs is as follows. The PMv and AIP are included in the model as the core nodes for grasping, as has been established in macaque studies [Luppino et al.,1999]. The DIFF > SAME contrast in the ANOVA revealed the pITG as well as PMv and AIP, suggesting that the selection of grip hand shape involves the pITG which locates in the ventral stream for object recognition [Grill‐Spector et al.,1998,2000]. We modeled connection between the AIP and pITG based on a recent macaque anatomical study in which tracer arrived at the inferotemporal cortex (TE, TEpd, TEm, and TEa) [Borra et al.,2008]. In addition to these three regions, we model the two nodes that take input of the visual signal (V1) and input of the experimental instruction (DLPFC). The V1 was selected since it serves as the common source of the dorsal and ventral visual streams, and also as the first cerebral gate to the visual input which spreads into the cerebral cortex. We used a focus in the V1 revealed in the effect of interest contrast. For simplicity of the models, we did not include the V2, V3/V3A, and the lateral intraparietal area (LIP) that may relay the dorsal signal from the V1 to AIP [Nakamura et al.,2001]. These omitted dorsal regions may also have communication with the ventral visual stream, but we consider that the interaction of the two visual streams can be represented, at least in part, by the interaction between the AIP and pITG.

The DLPFC activation was significant for the interaction in the ANOVA and also in the effect of interest contrast. We interpret the DLPFC activities may reflect the processing of the visual symbols that specify where and how to grip the object. Hence, we modeled the DLPFC as a node that receives the grip‐type instructions, namely the SAME/DIFF and the PRE/POW categorical signals. As for the connections, we modeled the connection of DLPFC‐AIP based on the recent anatomical study in macaque [Borra et al.,2008]. Although we do not have unequivocal experimental evidence that support a direct anatomical connection of the PMv‐pITG and DLPFC‐PMv [Rizzolatti and Luppino,2001], we incorporated these connections in the models and tested if they are likely by means of the random effect Bayesian model selection procedure [Penny et al.,2004]. All the connections were modeled as bidirectional.

Time series data extraction from the VOIs

First, we reanalyzed the functional data with new design matrices. New design matrices explicitly encoded the input and the two modulatory effects to conform to the neural state model in the DCM analysis. In the DCM analysis, the context‐dependent modulation of connections and driving input to the system are modeled explicitly and separately. In practice, the experimental blocks were regrouped into the input, and the two modulatory effects (GRIP‐CHANGE and TYPE). For the input, all second presentation blocks [POW2same, POW2diff, PRE2same, and PRE2diff conditions] were encoded with 1 and the first presentation [POW1 and PRE1] were 0. For the modulatory effect by GRIP‐CHANGE, DIFFERENT was encoded with 1 and SAME was with 0. For the modulatory effect by TYPE, POW as 1, and PRE as 0. The six motion parameters were also modeled as covariates of no interest in the design matrices. The effect of interest was calculated as the F contrast (SPM{F}) of the input and the two modulatory effects. The null hypothesis (H 0) of this effect of interest contrast is that the estimates of the three effects are all equal to zero, thus the rejection of the H 0 means at least one effect is significantly non zero.

Second, the time series data were extracted from the five VOIs that showed significant activation by this effect of interest F contrast. The individual VOIs were created as 6 mm radius spheres with centers that were the closest individual maxima to the group maxima identified in the ANOVA. The coordinates of group maxima (in MNI space) were taken from the effect of interest contrast in the ANOVA as follows: PMv (−42 8 31), AIP (−48 −37 46), DLPFC (−45 32 25), pITG (−48 −52 −8), and V1 (−9 −94 −8)). The coordinates of the AIP were selected as a local maximum that fall within the anatomical criteria of AIP, namely the junction of the intraparietal sulcus and the posterior central sulcus [Binkofski et al.,1998]. The coordinates of the pITG was obtained from the DIFF > SAME contrast in the ANOVA (Table II). The distance between the individual maxima and the group maxima were as follows: PMv (mean 0.83 mm, SD 1.4 mm), AIP (0.38, 1.0), DLPFC (0.89, 1.7), pITG (0.83, 1.5), and V1 (0,0). All the individual maxima were found within 6 mm from the group maxima. The F contrasts were thresholded with P < 0.05 without correction and the VOI time series data were extracted as the first principal component of the supra‐threshold voxels' time series data in the spherical VOI. The VOI time series were adjusted for the effect of interest; the signal change attributed to the motion parameters were subtracted from the time series data.

Table II.

Main effect of DIFF/SAME detected by the ANOVA for repetition effect

P (cluster) K (cluster) T (Peak voxel) MNI coordinates anatomical region
DIFF>SAME
0.001 127 6.23 −48 −52 −8 left pITGa
3.56 −45 −70 −8 left LOC
0.001 121 4.57 −30 −67 31 left pIPS
4.44 −33 −61 40 left IPS
3.95 −51 −34 43 left AIP
0.011 75 4.53 −42 8 31 left PMv
0.003 99 4.29 −36 −61 −23 left Cer
4.18 −30 −73 −14 left FG
3.98 −42 −73 −17 left LO
SAME>DIFF
0.026 61 5.32 15 −67 −8 Right LG
Open in a new tab

Note that the contrasts between the first and the second presentation (i.e., repetition effect; POW2diff‐POW1, POW2same‐POW1, PRE2diff‐PRE1, and PRE2same‐PRE1) were fed into the ANOVA.

Height threshold: T = 3.28 (P = 0.001 uncorrected), and extent threshold k = 50 voxels, P = 0.053 corrected.

a

This coordinate is used to define the group maxima with which we defined individual maxima to extract time series data for the DCM analysis. pITG posterior inferior temporal gyrus, LOC: lateral occipital cortex, pIPS: posterior intraparietal sulcus, IPS: intraparietal sulcus, AIP: anterior intraparietal area, PMv: ventral premotor area, Cer: cerebellum, FG: fusiform gyrus, LG: lingual gyrus.

Model estimation and comparison

A total of 28 DCMs were estimated and compared (Fig. 4A–C). In all models, the V1 receives the driving input, and the DLPFC received DIFF/SAME and the POW/PRE categorical signals. The connections PMv‐AIP, AIP‐V1, and pITG‐V1 were supposed to exist in all models (Fig. 4A). The connections DLPFC‐AIP, DLPFC‐PMv, and DLPFC‐pITG were allowed to have seven variations (Fig. 4B), and the connections PM‐pITG and AIP‐pITG were allowed four variations (Fig. 4C), thereby resulting a total of 28 models. All connections were modeled as bidirectional (i.e., PMv‐AIP means both PMv→AIP and AIP→PMv), and modulation by the factors GRIP‐CHANGE and TYPE was allowed on all connections except the self‐recursive ones. The all individual DCM estimates were submitted into the random effects Bayesian model selection procedure [Penny et al.,2004] separately for each session. The DCM estimates of connection strengths of the best model were averaged over sessions by the Bayesian approach and evaluated by one‐sample t tests with a statistical threshold of P < 0.05 with Bonferroni correction for multiple comparisons.

Figure 4.

Figure 4

Open in a new tab

DCM analysis. Top: Models. A: Fixed connections throughout all the models. The V1 receives the visual input, while the DLPFC receives the instruction that specifies the grip type. The bidirectional connections PMv‐AIP, V1‐AIP, and V1‐pITG were fixed to exist in all the models. B: The bidirectional connections between DLPFC‐ AIP, DLPFC‐ PMv, and DLPFC‐ pITG were allowed seven variations, namely [p q r] = [1 1 1],[1 1 0],[1 0 1],[0 1 1],[1 0 0],[0 1 0],or [0 0 1], where 0 means no connection is supposed and 1 means bidirectional connection is supposed. C: Similar to B, the bidirectional connection between pITG‐ AIP and pITG‐PMv had four variations, namely [s t] = [1 1],[1 0],[0 1],[0 0]. Combination of A–C makes a total of 28 models. Bottom: The best model in the Baysial model comparison. D. The full connection model was the best model. The one‐sample t tests showed significant connections in all the connections except AIP→DLPFC, AIP→V1, and PMv→DLPFC. Mean and SD of strength estimates and the result of one sample t test (P value) are given in Table IV. E. Significant modulation of the connection strength by the factor GRIP‐CHANGE. DIFFERENT grips compared with SAME grips in the second presentation increased the bottom‐up ventral visual stream (V1→pITG→PMv) and the interaction of the dorsal (AIP) and ventral (pITG) visiual areas. The prefrontal (DLPFC), the dorsal (AIP) and ventral (pITG) visual signals converge on the PMv, where the grip hand posture may be selected. Mean and SD of strength estimates and the result of the one sample t test (P value) are given in Table V. F. Significant modulation of the connection strength by the factor TYPE. Power grips had a stronger connection in AIP→PMv compared with precision grips. Mean and SD of strength estimates and the result of the one sample t test (P value) are given in Table VI.

RESULTS

Behavioral Data

Each movement of each subject was evaluated whether the hand posture followed the instruction symbols (precision or power grips) and whether the details such as aperture and wrist angle were changed from one trial to the next. Eleven of sixteen subjects performed the miming of grasping without errors. The mean error rate was 2.0% ± 3.3% SD.

MRI Data

First, we compared brain activation during the first presentation (PRE1 and POW1) against the baseline condition. Similar brain networks were active during the first grasp of objects by precision grip (PRE1) and power grip (POW1) (Fig. 2A,B). As expected, we found activation in the PMv, AIP, as well as the primary motor area (M1), dorsal premotor area (PMd), postocentral gyrus (PoCG), inferior parietal lobule (IPL), posterior inferior temporal gyrus (pITG), supplementary motor area (SMA), thalamus and cerebellar hemispheres (Cer). Second, the direct comparison revealed that the POW1 activation was higher than the PRE1 in the PMv, AIP, pITG, and other regions (Fig. 2C). The reverse contrast (PRE1 > POW1) did not detect significant activation. Third, we performed a within‐subject ANOVA on the contrast images for the REPETITION effects, namely POW2same‐POW1, POW2diff‐POW1, PRE2same‐PRE1, and PRE2diff‐PRE1. A significant interaction was found in the PMv, DLPFC, and AIP along with other cortices lining the intraparietal sulcus (Fig. 2E, Table I). The interaction reflects a signal increase exclusively in the PRE2diff condition compared to the signal decrease in the other three conditions (see Fig. 3). The main effect of GRIP‐CHANGE was found in the PMv, AIP, and pITG in a t contrast for DIFF > SAME (Fig. 2F,H left, Table II). The pITG cluster seems to be close to or to overlap with the lateral occipital complex (LOC) [Grill‐Spector et al.,1998,2000] and the fusiform gyrus [Buckner et al.,1998] that shows neural adaptation (i.e., repetition suppression) to object identity. Of note is that the pITG modulation is found only in the left hemisphere. As the left FG (−40 −52 −6) near to pITG (−48 −52 −8) is reported to show semantic adaptation [Koutstaal et al.,2001], the enhanced activity in the left pITG may indicate a change in the semantic processing of objects. For example, a photograph of a kettle (see Fig. 1) may activate the semantic memory of either the handle of the kettle or the handle on the lid rather than the kettle as a whole. The reverse contrast (DIFF < SAME) revealed the right lingual gyrus (V1/2) (Fig. 2H right, Table II). The main effect of TYPE (as a t contrast for PRE > POW) was found in the left LOC, bilateral dorsal visual stream, and the right basal ganglia (Fig. 2G, Table III). The POW > PRE contrast revealed no significant effect.

Figure 2.

Figure 2

Open in a new tab

Brain activation revealed by the contrasts for the first presentation (A‐C) against base line and the ANOVA for the difference Images (D‐H). A‐C. Brain activation by power grip and precision grip at the first presentation. A: POW1, B: PRE1, C: POW1 > PRE1. The reverse contrast, namely PRE1 > POW1, did not reveal significant activation. D‐H: ANOVA results. D, Effect of interest. E, Interaction of GRIP‐CHANGE and TYPE. F, DIFF > SAME. G, PRE > POW. H left, left pITG activation by the DIFF > SAME contrast. Right, right lingual gyrus activation by the SAME > DIFF contrast. All the maps were corrected P < 0.05. D is an F map, and others are t maps. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Table I.

Interaction detected by the ANOVA for repetition effect

P (cluster) K (cluster) T (Peak voxel) MNI coordinates Anatomical region
0 182 4.64 15 −55 −26 right Cer
4.44 3 −58 −17 vermis
0.011 75 4.61 27 −4 46 right PMd
0 256 4.57 33 −67 37 right pIPS
4.49 39 −52 40 right IPL
0 184 4.48 −45 8 28 left PMv
3.98 −45 29 28 left DLPFC
0 277 4.22 −48 −43 40 left IPL
4.01 −30 −70 37 left pIPS
Open in a new tab

Note the contrasts between the first and the second presentation (i.e., repetition effect; POW2diff‐POW1, POW2same‐POW1, PRE2diff‐PRE1, and PRE2same‐PRE1) were fed into the ANOVA. Interaction was calculated as a t contrast. Height threshold: T = 3.28 (P = 0.001 uncorrected), and extent threshold k = 50 voxels (P = 0.053 corrected for cluster level inference).

Cer: cerebellum, PMd: dorsal premotor area, pIPS: posterior intraparietal sulcus, IPL: inferior parietal lobule, PMv: ventral premotor area, DLPFC, dorsolateral prefrontal cortex, IPL: inferior parietal lobule, pIPS: posterior intra parietal sulcus. The tables were created by selecting up to three maxima in the clusters more than 4.0‐mm apart. When more than one maximum was found in the same anatomical region, only the one with the highest t value was shown.

Figure 3.

Figure 3

Open in a new tab

Mean VOI signals for the repetition effect. The BOLD signal difference between the first and the second presentation (i.e., BOLD signals in the second presentation minus those in the first). BOLD signal enhancement is found only in the PRE2diff condition, and other conditions show signal suppression. Signals are represented in an arbitrary unit, but common to all VOIs.

Table III.

Main effect of POW/PRE detected by the ANOVA for repetition effect

P (cluster) K (cluster) T (Peak voxel) MNI coordinates Anatomical region
PRE>POW
0 684 7.73 −12 −94 −5 left V1/2
6.05 −30 −55 −11 left FG
5.58 −39 −76 −11 left LO
0 774 6.86 18 −76 −17 right LG
6.04 27 −64 −14 right FG
5.8 12 −76 −5 right V1/2
0 179 4.29 30 −73 34 right pIPS
4.08 27 −61 55 right SPL
0.031 58 4.15 3 14 −5 right CH
3.75 9 2 −2 right GP
POW>PRE N.S.
Open in a new tab

Note that the contrasts between the first and the second presentation (i.e., repetition effect; POW2diff‐POW1, POW2same‐POW1, PRE2diff‐PRE1, and PRE2same‐PRE1) were fed into the ANOVA.

Height threshold: T = 3.28 (P = 0.001 uncorrected), and extent threshold k = 50 voxels, P = 0.053 corrected.

FG: fusiform gyrus, LOC: lateral occipital cortex, LG: lingual gyrus, FG: fusiform gyrus, V1/2: visual area 1 or 2, pIPS: posterior intraparietal sulcus, SPL: superior parietal lobule, CH, caudate head, GP: globus pallidus.

DCM Results

The Bayesian model selection procedure was applied for three sessions separately. The three results all indicated that the model that allows all the connections is the best model (see Fig. 4). The model expected probabilities of the winning model were 6.75, 6.94, and 5.87 × 10−2 in the Session 1–3, while those of other 27 models were less than 3.75 × 10−2. The model exceedance probabilities of the winning model were 1.30, 1.41, and 1.08 × 10−1, while those of other 27 models were less than 4.14 × 10−2. The one‐sample t tests on the connection strength revealed all the intrinsic connection strengths are significant (Fig. 4D, Table IV) except AIP→DLPFC, AIP→V1, and PMv→DLPFC. The same statistical tests on the modulatory effect by GRIP‐CHANGE found that DIFFERENT relative to SAME augments the convergence of the causal influences on the PMv from the AIP, DLPFC, and pITG (Fig. 4E, Table V). It also increases the interaction between the ventral (pITG) and dorsal (AIP) visual streams. The t tests on the modulatory effect by TYPE revealed that the power grip, relative to the precision grip, increased the causal influence from the AIP to PMv (Fig. 4F, Table VI). Since our current knowledge of anatomical connectivity of the human brain is limited, the best DCM model may not exclusively be comprised of direct connections. Rather, it suggests the most likely causal influence flow possibly including indirect connections via hidden (i.e., not modeled, such as DLPFC→area 24→F6→F5 (PMv) [Rizzolatti and Luppino,2001] regions. Even after taking the uncertainty in the directness of connections into account, however, we can summarize that the present DCM results imply the interaction of the dorsal and ventral visual streams and the convergence of causal influences on the PMv.

Table IV.

One‐sample t test on connection strength parameters estimated in the best model (Model 1) by DCM analysis: Intrinsic connection

To/from DLPFC PMv AIP pITG V1
DLPFC −1.000(0.000) P = 0.00e+00 0.044(0.060) P = 1.03e‐02 0.082(0.092) P = 2.62e‐03 0.066(0.071) P = 2.25e‐03
PMv 0.112(0.055) P = 7.05e‐07 ‐1.000(0.000) P = 0.00e+00 0.178(0.087) P = 6.20e‐07 0.131(0.068) P = 1.45e‐06
AIP 0.101(0.069) P = 3.14e‐05 0.093(0.069) P = 6.70e‐05 −1.000(0.000) P = 0.00e+00 0.134(0.075) P = 3.43e‐06 0.253(0.117) P = 3.44e‐07
pITG 0.096(0.072) P = 8.55e‐05 0.072(0.066) P = 5.04e‐04 0.119(0.068) P = 4.51e‐06 −1.000(0.000) P = 0.00e+00 0.152(0.058) P = 2.53e‐08
V1 0.051(0.065) P = 6.09e‐03 0.055(0.057) P = 1.53e‐03 −1.000(0.000) P = 0.00e+00
Open in a new tab

Threshold is P > 0.0024 (Bonferroni correction of P < 0.05 for 21 comparisons). Mean (SD) and probability are given for each connection. Significant connections are printed in bold. The empty cells are for connections that were not assumed.

Table V.

One‐sample t test on connection strength parameters estimated in the best model (Model 1) by DCM analysis: Modulation by GRIP‐CHANGE

To/from DLPFC PMv AIP pITG V1
DLPFC −0.001(0.069) P = 9.67e‐01 −0.006(0.070) P = 7.56e‐01 0.004(0.066) P = 8.01e‐01
PMv 0.067(0.039) P = 4.97e‐06 0.087(0.056) P = 1.46e‐05 0.071(0.038) P = 2.16e‐06
AIP 0.045(0.068) P = 1.84e‐02 0.027(0.056) P = 7.87e‐02 0.043(0.045) P = 1.77e‐03 0.055(0.078) P = 1.23e‐02
pITG 0.044(0.051) P = 3.46e‐03 0.023(0.041) P = 3.97e‐02 0.040(0.038) P = 7.31e‐04 0.068(0.056) P = 2.03e‐04
V1 0.005(0.064) P = 7.83e‐01 0.011(0.063) P = 4.76e‐01
Open in a new tab

Threshold is P > 0.0031 (Bonferroni correction of P < 0.05 for 16 comparisons). Mean (SD) and probability are given for each connection. Significant connections are printed in bold.

Table VI.

One‐sample t test on connection strength parameters estimated in the best (Model 1) by DCM analysis: Modulation by TYPE

To/from DLPFC PMv AIP pITG V1
DLPFC 0.002(0.050) P = 9.01e‐01 −0.002(0.042) P = 8.60e‐01 −0.004(0.054) P = 7.52e‐01
PMv 0.039(0.052) P = 8.91e‐03 0.055(0.059) P = 1.98e‐03 0.036(0.059) P = 2.71e‐02
AIP 0.028(0.056) P = 6.13e‐02 0.030(0.044) P = 1.40e‐02 0.028(0.042) P = 1.68e‐02 0.023(0.083) P = 2.90e‐01
pITG 0.030(0.057) P = 4.90e‐02 0.025(0.048) P = 5.84e‐02 0.026(0.050) P = 5.70e‐02 0.005(0.062) P = 7.35e‐01
V1 −0.027(0.061) P = 1.02e‐01 −0.015(0.039) P = 1.39e‐01
Open in a new tab

Threshold is P > 0.0031 (Bonferroni correction of P < 0.05 for 16 comparisons). Mean (SD) and probability are given for each connection. Significant connections are printed in bold.

DISCUSSION

Interaction of Dorsal and Ventral Visual Streams, and Grip Selection at the PMv

We consider that grip selection involves the PMv, AIP, and pITG, which showed higher activities in the DIFFERENT conditions than in the SAME conditions (i.e., the main effect of GRIP‐CHANGE). We attribute the BOLD signal increase to the cost of overriding the previous neural computation in each region. In addition to these changes in the regional signal amplitude, the DCM results further suggest a dynamic change in the network in response to the increased selection demands (Fig. 4E). First, it increases the bottom‐up neural activities in the course of V1→pITG→PMv. This suggests the altered grip may require an update in object recognition, for example from the handle of the kettle to the handle on the lid (see Fig. 1). Second, the interaction increases between the dorsal (AIP) and the ventral (pITG) visual areas. Computation of affordable hand movements in the dorsal visual pathway may be biased by the attended part (e.g., handle of the kettle or the handle on the lid) of the object. Lastly, the neural information representing the task instruction (DLPFC), possible hand shapes (AIP), and object recognition (pITG) converges on the PMv, where the selection of grip seems to happen. The grip selection may be achieved by the integration of distributed heteromodal information.

Grasping Real Objects and Semantic Information Retrieval

Several lines of research have suggested that grasping does not need recognition of the objects. Lesion studies reported a double dissociation between grasping and object recognition due to occipitotempolal [Goodale et al.,1991] and parietal [Jakobson et al.,1991] damage. Neuroimaging techniques also revealed that grasping activates the dorsal visual pathway as opposed to the ventral visual pathway responsible for object recognition [Culham et al.,2003; James et al.,2003]. However, we often grasp an identified object as the first action of meaningful goal‐directed behavior, and it is indeed hard to exclude the involvement of the semantic memory retrieval process from grasping objects of everyday use. For example, one will be surprised if one picks up a book which is extremely heavy or light. This suggests that the plausible weight is retrieved from long‐term semantic memory through visual identification of the object in advance of the execution of the grasping/lifting movement [Gordon et al.,1993]. Similarly, how to grasp is also influenced by the knowledge and the goal of action. One will take the handle of a cup after pouring hot coffee into it, but one may grab an empty cup anywhere if one fetches it for washing. The present study suggests interaction between the grip formation and the object recognition: The dorsal (AIP) and the ventral visual streams (pITG) have significant intrinsic connections and there is also a significant task dependent modulation of the connection strength between them, which complements preceding anatomical study [Borra et al.,2008].

Asymmetric Effects of Prior Grip Type

Although we predicted signal increases in regions involved in grip selection when the subjects made a different grip than the first one, POW2diff condition did not show a signal increase. These asymmetric effects between the grip types might be explained by the difference in the computational load between the two grips. The greater activation in POW1 than PRE1 may indicate that POW1 require more computation than PRE1. The higher computational load in power grip may be accounted for by greater degree of freedom in forming the grip hand shape than in the precision grip. For example, more hand muscles must be controlled in power grip. When the first task (POW1) was different and computationally heavier than the second one (PRE2), the neural cost of overriding the previous response may be higher.

Laterality

The present study found brain activation predominantly in the left hemisphere. This may be partly explained by the use of the right hand, but clinical observations suggest the significant lateralization of the ability to combine the visual recognition of objects with the associated action. The miming of tool use is selectively disturbed with left brain damage [Goldenberg et al.,2003], especially in inferior frontal damage patients [Goldenberg et al.,2007]. Furthermore, split‐brain patients have difficulty in the pantomiming of object use only when they use their left hands [Lausberg et al.,2003], suggesting the critical importance of the left hemisphere for this. The leftward lateralized activation is also remarkable in the planning of tool‐use gestures even using the left hand [Johnson‐Frey et al.,2005]. Thus, converging evidence suggests that the left hemisphere has a dominance in controlled access to semantic memory not only in linguistic but also in motoric domains.

Pantomiming and Real Movement

Although we identified regions that were consistently found in real grasping movement studies such as the PMv, AIP, and pITG, we need to consider the possibility that the use of pantomiming can limit the generalizability to real grasping. Studies on brain‐damaged patients have clearly demonstrated the double dissociation between real and pantomimed grasping. Visual form agnosia patients who have lesions in the ventral visual pathway can grasp 3D objects but fail to pantomime the same action [Goodale et al.,1991,1994]. By contrast, patients with lesions to the dorsal visual pathway, as in optic ataxia, fail in real grasping, but perform the same movement better when pantomiming [Milner et al.,2001]. Furthermore, dissociation is also evidenced in the kinematics of the actual and fake grasping movements in brain‐damaged patients [Hermsdorfer et al.,2006; Laimgruber et al.,2005] and in normals [Westwood et al.,2000].

Motivated by this line of evidence, a fMRI study critically examined if real and pantomimed grasping differed in terms of brain activation [Kroliczak et al.,2007]. This study found that pantomimed grasping activated the same regions at a consistently lower level than real grasping did, but this quantitative relation is reversed in the right temporal region [Kroliczak et al.,2007]. This fMRI finding, together with clinical observations, opens up the possibility that there are different neural mechanisms for real and pantomimed grasping. Therefore, further research is necessary to ascertain if the present findings are generalizable to the neural mechanisms of real grasping movements.

Instead, the present study provides information about the neural system dynamics for hand shape formation toward meaningful visual objects. First, we identified the brain network dynamics in response to the increased selection demands. Indeed, the present study replicated the finding of a PMv activation in response to increased selection demands reported in a word generation study [Thompson‐Schill et al.,1999], suggesting a domain‐general mechanism in selecting a response from competing candidates. The present study further suggests that the enhanced PMv activation reflects the increased causal influence to the region from posterior regions such as parietal and inferior temporal cortices, as well as from the prefrontal cortex, which specifies the current task. Secondly, the use of daily (i.e., meaningful) objects enabled us to detect the interaction between the dorsal (“where”) and the ventral (“what”) visual streams [Ungerleider and Haxby,1994] in the formation of the hand motor response associated with the meanings of the object. Although the present results may not be directly generalizable to real grasping, it suggests general mechanisms in the selection of motor response and an interaction between the dorsal and ventral visual streams in the context of hand shape formation toward meaningful visual objects.

Acknowledgements

The authors are grateful to Minako Kawai and Mihoko Naruo for the evaluation of the behavioral performances, to Rosie Wallis for English correction, to Kerstin Flake and Andrea Gast‐Sandmann for the graphics, to Kimihiro Nakamura for helpful comments on the earlier version of manuscript, and finally to Angela D. Friederici for support. Special thanks to Akihiro Sasaki and Takanori Kochiyama for guiding the DCM analysis.

†

This study was performed in the Hamano Life Science Research Foundation Ogawa Laboratories for Brain Function Research when the CA was affiliated to the Department of Speech Physiology, Graduate School of Medicine, University of Tokyo.

REFERENCES

  1. Ashburner J ( 2007): A fast diffeomorphic image registration algorithm. Neuroimage 38: 95–113. [DOI] [PubMed] [Google Scholar]
  2. Binkofski F, Dohle C, Posse S, Stephan KM, Hefter H, Seitz RJ, Freund HJ ( 1998): Human anterior intraparietal area subserves prehension: A combined lesion and functional MRI activation study. Neurology 50: 1253–1259. [DOI] [PubMed] [Google Scholar]
  3. Borra E, Belmalih A, Calzavara R, Gerbella M, Murata A, Rozzi S, Luppino G ( 2008): Cortical connections of the macaque anterior intraparietal (AIP) area. Cereb Cortex 18: 1094–1111. [DOI] [PubMed] [Google Scholar]
  4. Buckner RL, Goodman J, Burock M, Rotte M, Koutstaal W, Schacter D, Rosen B, Dale AM ( 1998): Functional‐anatomic correlates of object priming in humans revealed by rapid presentation event‐related fMRI. Neuron 20: 285–296. [DOI] [PubMed] [Google Scholar]
  5. Castiello U, Begliomini C ( 2008): The cortical control of visually guided grasping. Neuroscientist 14: 157–170. [DOI] [PubMed] [Google Scholar]
  6. Chong TT, Cunnington R, Williams MA, Kanwisher N, Mattingley JB ( 2008): fMRI adaptation reveals mirror neurons in human inferior parietal cortex. Curr Biol 18: 1576–1580. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Cisek P, Kalaska JF. ( 2010): Neural mechanisms for interacting with a world full of action choices. Annu Rev Neurosci 33: 269–298. [DOI] [PubMed] [Google Scholar]
  8. Culham JC, Danckert SL, DeSouza JF, Gati JS, Menon RS, Goodale MA ( 2003): Visually guided grasping produces fMRI activation in dorsal but not ventral stream brain areas. Exp Brain Res 153: 180–189. [DOI] [PubMed] [Google Scholar]
  9. Ehrsson HH, Fagergren A, Jonsson T, Westling G, Johansson RS, Forssberg H ( 2000): Cortical activity in precision‐ versus power‐grip tasks: An fMRI study. J Neurophysiol 83: 528–536. [DOI] [PubMed] [Google Scholar]
  10. Fagg AH, Arbib MA ( 1998): Modeling parietal‐premotor interactions in primate control of grasping. Neural Netw 11: 1277–1303. [DOI] [PubMed] [Google Scholar]
  11. Fogassi L, Gallese V, Buccino G, Craighero L, Fadiga L, Rizzolatti G ( 2001): Cortical mechanism for the visual guidance of hand grasping movements in the monkey: A reversible inactivation study. Brain 124( Part 3): 571–586. [DOI] [PubMed] [Google Scholar]
  12. Friston KJ, Harrison L, Penny W ( 2003): Dynamic causal modelling. Neuroimage 19: 1273–1302. [DOI] [PubMed] [Google Scholar]
  13. 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. [DOI] [PubMed] [Google Scholar]
  14. Goldenberg G, Hartmann K, Schlott I ( 2003): Defective pantomime of object use in left brain damage: Apraxia or asymbolia? Neuropsychologia 41: 1565–1573. [DOI] [PubMed] [Google Scholar]
  15. Goldenberg G, Hermsdorfer J, Glindemann R, Rorden C, Karnath HO ( 2007): Pantomime of tool use depends on integrity of left inferior frontal cortex. Cereb Cortex 17: 2769–2776. [DOI] [PubMed] [Google Scholar]
  16. Goodale MA, Milner AD ( 1992): Separate visual pathways for perception and action. Trends Neurosci 15: 20–25. [DOI] [PubMed] [Google Scholar]
  17. Goodale MA, Milner AD, Jakobson LS, Carey DP ( 1991): A neurological dissociation between perceiving objects and grasping them. Nature 349: 154–156. [DOI] [PubMed] [Google Scholar]
  18. Goodale MA, Jakobson LS, Keillor JM ( 1994): Differences in the visual control of pantomimed and natural grasping movements. Neuropsychologia 32: 1159–1178. [DOI] [PubMed] [Google Scholar]
  19. Gordon AM, Westling G, Cole KJ, Johansson RS ( 1993): Memory representations underlying motor commands used during manipulation of common and novel objects. J Neurophysiol 69: 1789–1796. [DOI] [PubMed] [Google Scholar]
  20. Grill‐Spector K, Kushnir T, Edelman S, Itzchak Y, Malach R ( 1998): Cue‐invariant activation in object‐related areas of the human occipital lobe. Neuron 21: 191–202. [DOI] [PubMed] [Google Scholar]
  21. Grill‐Spector K, Kushnir T, Hendler T, Malach R ( 2000): The dynamics of object‐selective activation correlate with recognition performance in humans. Nat Neurosci 3: 837–843. [DOI] [PubMed] [Google Scholar]
  22. Grill‐Spector K, Henson R, Martin A ( 2006): Repetition and the brain: Neural models of stimulus‐specific effects. Trends Cogn Sci 10: 14–23. [DOI] [PubMed] [Google Scholar]
  23. Grol MJ, Majdandzic J, Stephan KE, Verhagen L, Dijkerman HC, Bekkering H, Verstraten FA, Toni I ( 2007): Parieto‐frontal connectivity during visually guided grasping. J Neurosci 27: 11877–11887. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Hamilton AF, Grafton ST ( 2006): Goal representation in human anterior intraparietal sulcus. J Neurosci 26: 1133–1137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Hermsdorfer J, Hentze S, Goldenberg G ( 2006): Spatial and kinematic features of apraxic movement depend on the mode of execution. Neuropsychologia 44: 1642–1652. [DOI] [PubMed] [Google Scholar]
  26. Jakobson LS, Archibald YM, Carey DP, Goodale MA ( 1991): A kinematic analysis of reaching and grasping movements in a patient recovering from optic ataxia. Neuropsychologia 29: 803–809. [DOI] [PubMed] [Google Scholar]
  27. James TW, Culham J, Humphrey GK, Milner AD, Goodale MA ( 2003): Ventral occipital lesions impair object recognition but not object‐directed grasping: An fMRI study. Brain 126: 2463–2475. [DOI] [PubMed] [Google Scholar]
  28. Johnson‐Frey SH, Newman‐Norlund R, Grafton ST ( 2005): A distributed left hemisphere network active during planning of everyday tool use skills. Cereb Cortex 15: 681–695. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kilner JM, Neal A, Weiskopf N, Friston KJ, Frith CD ( 2009): Evidence of mirror neurons in human inferior frontal gyrus. J Neurosci 29: 10153–10159. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Koutstaal W, Wagner AD, Rotte M, Maril A, Buckner RL, Schacter DL ( 2001): Perceptual specificity in visual object priming: Functional magnetic resonance imaging evidence for a laterality difference in fusiform cortex. Neuropsychologia 39: 184–199. [DOI] [PubMed] [Google Scholar]
  31. Kroliczak G, Cavina‐Pratesi C, Goodman DA, Culham JC ( 2007): What does the brain do when you fake it? An FMRI study of pantomimed and real grasping. J Neurophysiol 97: 2410–2422. [DOI] [PubMed] [Google Scholar]
  32. Laimgruber K, Goldenberg G, Hermsdorfer J ( 2005): Manual and hemispheric asymmetries in the execution of actual and pantomimed prehension. Neuropsychologia 43: 682–692. [DOI] [PubMed] [Google Scholar]
  33. Lausberg H, Cruz RF, Kita S, Zaidel E, Ptito A ( 2003): Pantomime to visual presentation of objects: Left hand dyspraxia in patients with complete callosotomy. Brain 126( Part 2): 343–360. [DOI] [PubMed] [Google Scholar]
  34. Luppino G, Murata A, Govoni P, Matelli M ( 1999): Largely segregated parietofrontal connections linking rostral intraparietal cortex (areas AIP and VIP) and the ventral premotor cortex (areas F5 and F4). Exp Brain Res 128: 181–187. [DOI] [PubMed] [Google Scholar]
  35. Milner AD, Dijkerman HC, Pisella L, McIntosh RD, Tilikete C, Vighetto A, Rossetti Y ( 2001): Grasping the past. Delay can improve visuomotor performance. Curr Biol 11: 1896–1901. [DOI] [PubMed] [Google Scholar]
  36. Murata A, Fadiga L, Fogassi L, Gallese V, Raos V, Rizzolatti G ( 1997): Object representation in the ventral premotor cortex (area F5) of the monkey. J Neurophysiol 78: 2226–2230. [DOI] [PubMed] [Google Scholar]
  37. Murata A, Gallese V, Luppino G, Kaseda M, Sakata H ( 2000): Selectivity for the shape, size, and orientation of objects for grasping in neurons of monkey parietal area AIP. J Neurophysiol 83: 2580–2601. [DOI] [PubMed] [Google Scholar]
  38. Nakamura H, Kuroda T, Wakita M, Kusunoki M, Kato A, Mikami A, Sakata H, Itoh K ( 2001): From three‐dimensional space vision to prehensile hand movements: The lateral intraparietal area links the area V3A and the anterior intraparietal area in macaques. J Neurosci 21: 8174–8187. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Penny WD, Stephan KE, Mechelli A, Friston KJ ( 2004): Comparing dynamic causal models. Neuroimage 22: 1157–1172. [DOI] [PubMed] [Google Scholar]
  40. Raos V, Umilta MA, Murata A, Fogassi L, Gallese V ( 2006): Functional properties of grasping‐related neurons in the ventral premotor area F5 of the macaque monkey. J Neurophysiol 95: 709–729. [DOI] [PubMed] [Google Scholar]
  41. Rizzolatti G, Luppino G ( 2001): The cortical motor system. Neuron 31: 889–901. [DOI] [PubMed] [Google Scholar]
  42. Rizzolatti G, Camarda R, Fogassi L, Gentilucci M, Luppino G, Matelli M ( 1988): Functional organization of inferior area 6 in the macaque monkey. II. Area F5 and the control of distal movements. Exp Brain Res 71: 491–507. [DOI] [PubMed] [Google Scholar]
  43. Sakata H, Taira M, Murata A, Mine S ( 1995): Neural mechanisms of visual guidance of hand action in the parietal cortex of the monkey. Cereb Cortex 5: 429–438. [DOI] [PubMed] [Google Scholar]
  44. Shmuelof L, Zohary E ( 2005): Dissociation between ventral and dorsal fMRI activation during object and action recognition. Neuron 47: 457–470. [DOI] [PubMed] [Google Scholar]
  45. Thompson‐Schill SL, D'Esposito M, Aguirre GK, Farah MJ ( 1997): Role of left inferior prefrontal cortex in retrieval of semantic knowledge: A reevaluation. Proc Natl Acad Sci USA 94: 14792–14797. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Thompson‐Schill SL, D'Esposito M, Kan IP ( 1999): Effects of repetition and competition on activity in left prefrontal cortex during word generation. Neuron 23: 513–522. [DOI] [PubMed] [Google Scholar]
  47. Ungerleider LG, Haxby JV ( 1994): “What” and “where” in the human brain. Curr Opin Neurobiol 4: 157–165. [DOI] [PubMed] [Google Scholar]
  48. Verhagen L, Dijkerman HC, Grol MJ, Toni I ( 2008): Perceptuo‐motor interactions during prehension movements. J Neurosci 28: 4726–4735. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Westwood DA, Chapman CD, Roy EA ( 2000): Pantomimed actions may be controlled by the ventral visual stream. Exp Brain Res 130: 545–548. [DOI] [PubMed] [Google Scholar]

Articles from Human Brain Mapping are provided here courtesy of Wiley

RESOURCES