Activation of the Dorsal Premotor Cortex and Pre-Supplementary Motor Area of Humans During an Auditory Conditional Motor Task

RAPID COMMUNICATION Activation of the Dorsal Premotor Cortex and Pre-Supplementary Motor Area of Humans During an Auditory Conditional Motor Task KIYOSHI KURATA, 1 TOSHIAKI TSUJI, 3 SATOSHI NARAKI, 3 MORIO SEINO, 3 AND YOSHINAO ABE 2 1 Department of Physiology and 2 Department of Radiology, Hirosaki University School of Medicine; and 3 Division of Radiology, Hirosaki University Hospital, Hirosaki 036-8562, Japan Received 4 January 2000; accepted in final form 16 May 2000 Kurata, Kiyoshi, Toshiaki Tsuji, Satoshi Naraki, Morio Seino, and Yoshinao Abe. Activation of the dorsal premotor cortex and pre-supplementary motor area of humans during an auditory conditional motor task. J Neurophysiol 84: 1667 1672, 2000. Using functional magnetic resonance imaging (fmri), we measured regional blood flow to examine which motor areas of the human cerebral cortex are preferentially involved in an auditory conditional motor behavior. As a conditional motor task, randomly selected 330 or 660 Hz tones were presented to the subjects every 1.0 s. The low and high tones indicated that the subjects should initiate three successive opposition movements by tapping together the right thumb and index finger or the right thumb and little finger, respectively. As a control task, the same subjects were asked to alternate the two opposition movements, in response to randomly selected tones that were presented at the same frequencies. Between the two tasks, MRI images were also scanned in the resting state while the tones were presented in the same way. Comparing the images during each of the two tasks with images during the resting state, it was observed that several frontal motor areas, including the primary motor cortex, dorsal premotor cortex (PMd), supplementary motor area (SMA), and pre-sma, were activated. However, preferential activation during the conditional motor task was observed only in the PMd and pre-sma of the subjects left (contralateral) frontal cortex. The PMd has been thought to play an important role in transforming conditional as well as spatial visual cues into corresponding motor responses, but our results suggest that the PMd along with the pre-sma are the sites where more general and extensive sensorimotor integration takes place. INTRODUCTION In a conditional motor behavior, desired motor responses are indicated by arbitrary sensory cues, such as colors, which are not primarily associated with the responses. Such behavior is extremely important in humans, because our actions are frequently instructed and guided by arbitrary cues, like traffic signals or language. In monkeys, it has been reported that the dorsal premotor cortex (PMd) is essential for performing conditional motor behavior (Kurata and Hoffman 1994; Kurata and Hoshi 1999; Mitz et al. 1991; Passingham 1985a,b; Petrides 1987). The ventral premotor cortex (PMv) plays only a minor role in the behavior but contributes to prism adaptation while reaching toward visuospatial targets (Kurata and Hoshi 1999). As in monkeys, positron emission tomography (PET) studies revealed that the PMd of humans is activated during learning and performance of conditional motor behavior (Deiber et al. 1997; Grafton et al. 1998; Hazeltine et al. 1997). Address reprint requests to K. Kurata (E-mail: kuratak@cc.hirosaki-u.ac.jp). In those studies of humans and monkeys, visual stimuli were used as the conditional cues. However, it is likely that cues with a nonvisual modality, namely auditory cues, also guide conditional motor behavior. Supporting this view, the activity of monkeys PMd neurons was reported to show sustained changes during the preparation period for forthcoming movements following auditory instruction cues (Kurata 1993; Weinrich and Wise 1982; Wise et al. 1996). In humans, patients with premotor (PM) lesions have deficits in conditional association of not only visual, but also tactile and auditory stimuli with movements (Halsband and Freund 1990). Furthermore, the PMd is activated when spatially (but not conditionally) presented auditory and visual cues indicate the laterality of hand movements (Iacoboni et al. 1998). If the PMd is activated by arbitrary auditory cues for conditional motor behavior, this indicates that the area may play a more extensive role in sensorimotor integration than has previously been considered. In addition to the PMd, we are particularly interested in the presupplementary motor area (pre-sma) of humans, because this area is also activated during learning of conditional visuomotor behavior (Deiber et al. 1997) and visuomotor associations (Sakai et al. 1999). Thus, the aim of this study was to measure the regional cerebral blood flow (rcbf) to examine whether the PMd and pre-sma of humans play a role in conditional motor behavior with arbitrary auditory cues. We therefore compared functional magnetic resonance imaging (fmri) scanned while performing two tasks; one required conditional association of auditory cues and appropriate motor responses, and the other combined the same movements and the same auditory cues, but without the conditional association. METHODS Five healthy right-handed volunteers (3 males and 2 females, 21 26 years old) participated in this study. None had any history of neurological or psychiatric disorders, and all gave informed consent before the experiment. They were trained to perform an auditory conditional motor task shortly before the scan. In the task, 330 or 660 Hz tones (100-ms duration) were randomly selected and presented to the subjects every 1.0 s through earphones with a sound pressure level equal to or greater than the maximal MRI noise level. The low and high tones were the signals to initiate three successive opposition move- The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. www.jn.physiology.org 0022-3077/00 $5.00 Copyright 2000 The American Physiological Society 1667

1668 K. KURATA, T. TSUJI, S. NARAKI, M. SEINO, AND Y. ABE ments of the right thumb and index finger or the right thumb and little finger, respectively (Task 1). As a control task (Task 2), the same subjects were asked to alternate the two opposition movements in response to randomly selected tones presented at the same interval. In our preliminary fmri study, it was confirmed that cortical motor areas, including the PMd and pre-sma, were activated by the rate and number of taps. Each subject was naïve to Tasks 1 and 2 and practiced the tasks for about 3 5 min each immediately before the MRI scan. In the experimental session, the two tasks were alternated for 30 s (30 trials), with an intervening rest period (Rest) of 30 s (30 tone signals). In the Rest condition, the same tones were presented as in the two tasks. Thus, one session consisted of four epochs (Task 1, Rest, Task 2, and Rest); six sessions were repeated in succession. Throughout the six sessions, fmri scans were taken continuously with a whole-body 1.5 Tesla scanner (Siemens Magnetom Vision 4). A video camera was used to monitor the subject s hand movements during the scan. Straps around the forehead and jaw were used to minimize head motion. Before obtaining the fmri data, three sagittal T1-weighted images around the midline and 10 horizontal T1-weighted images parallel to the line between the anterior and posterior commissures were scanned [repetition time (TR) 300 ms, echo time (TE) 14 ms, flip angle (FA) 90, slice thickness 3.0 mm, slice gap 0.3 mm, imaging matrix 256 256, and field of view (FOV) 256 256 mm] to identify the cortical areas with reference to the central and precentral sulci, and a vertical line traversing the anterior commissure (VCA line) in Talairach coordinates (Talairach and Tournoux 1988). Five slices from the 10 T1 images were used to obtain fmri images of the primary motor cortex (MI), the primary somatosensory cortex (SI), PMd, SMA, and pre-sma (see Fig. 1). After identifying the cortical sulci in the T1 images, each functional image was obtained using a single-shot echo-planar imaging (EPI) sequence. Each set of functional images consisted of five axial T2*-weighted images (TE 66 ms, FA 90, slice thickness 3.0 mm, slice gap 0.3 mm, an imaging matrix 128 128, and FOV 256 256 mm). Since a single-shot EPI sequence was used in each scan, there was no TR. The acquisition time of a five-image set was 0.82 s. An image set was acquired every 3 s throughout the entire scan, so that 10 sets were obtained in each 30 s epoch making up Task 1, Task 2, and Rest. Thus, 240 image sets were obtained for each subject during the six sessions, each set consisting of four epochs (Task 1, Rest, Task 2, and Rest). No event-related scans were attempted, so the interval between the cue and the timing of the scan was not synchronized in this study. After scanning the functional images, 50 T1-weighted images of the whole brain were collected (TR 300 ms, TE 14 ms, FA 90, slice thickness 3.0 mm, slice gap 0.3 mm, imaging matrix 256 256, and FOV 256 256 mm). Data were analyzed on a single subject basis. Using SPM96 for Windows NT (Wellcome Department of Cognitive Neurology, London, UK) with Matlab for Windows (v. 5.3, MathWorks, Sherborn, MA), fmri time series data were smoothed (full width at halfmaximum 4.00 mm) and analyzed. No spatial normalization of the data were performed. The first five scans were excluded from the analysis because of the nonequilibrium state of magnetization. Constructing three delayed box-car functions (delay 6 s), Task 1 versus Task 2, Task 1 versus Rest, and Task 2 versus Rest were compared using linear contrasts of [1, 0, 1], [1, 1, 0], and [0, 1, 1] for [Task 1, Rest, Task 2], respectively. In the analysis, data from the two Rest conditions were combined, and we applied a high-pass filter (cutoff period 120 s, equivalent to the time for each session) with temporal smoothing, no covariate of no interest, and no global normalization. We also examined whether there was a statistically significant difference between the Rest conditions after Task 1 and that after Task 2 using a contrast of [1, 1] for [Rest after Task 1, Rest after Task 2] in SPM96. To focus on significant activation in voxels, we measured Z scores with the unit normal distribution, and created {Z} maps consisting of the voxels with a threshold at P 0.005 (corrected for multiple comparisons) throughout the analysis. The activated foci were located by superimposing SPM{Z} onto the T1 images. The activated areas in the mesial cortex immediately caudal and rostral to the VCA line and dorsal to the cingulate sulcus are referred to as the SMA and pre-sma, respectively (Hikosaka et al. 1996; Vorobiev et al. 1998). The x, y, and z coordinates of the activated voxel in each region were also obtained using SPM96. RESULTS During scans, the subjects performed Task 1 correctly in more than 98% of the trials. Selection errors were only made in 2% of the trials; these occurred when the scanner made a loud noise that coincided with the auditory cues. No omission errors were observed during Task 1. The subjects did not make any mistakes during Task 2. Figure 1 shows the brain regions of a representative subject that were significantly activated during Task 1 versus Rest, Task 2 versus Rest, and Task 1 versus Task 2. During Task 1 versus Rest, areas immediately rostral and caudal to the left central sulcus were activated, corresponding to the MI and SI, respectively. The left activated region continued to the dorsal Brodmann s area 6 in the precentral gyrus, and this area was regarded as the PMd (Freund 1991; Grafton et al. 1998; Iacoboni et al. 1998; Wise et al. 1991). On the left mesial cortex (bottom left of Fig. 1), the left SMA, and pre- SMA were activated (see METHODS for definition). The right pre-sma was also activated. Minor activation was observed in the PMd and MI of the right hemisphere, but there were no activated areas in the prefrontal cortex. Table 1 shows the average location of constantly activated areas in the frontal cortex of the five subjects, and the average number of pixels in each area. Figure 1 and Table 1 indicate that the distribution and number of activated voxels in each area were similar when comparing Task 1 versus Rest or Task 2 versus Rest. When Task 1 and Task 2 were compared, however, preferentially activated regions were observed constantly in the left PMd and pre-sma contralateral to the working hand (Fig. 1). In contrast, no preferentially activated regions were observed in the MI, SI, or SMA. Figure 2 shows averaged signals in the region of interest (ROI) within the left PMd during the three conditions. In the ROI, the signals observed were always greater in Task 1 than in Task 2. The signals observed in both Tasks 1 and 2 were significantly greater than those during Rest. Similar changes were observed in the pre-sma. Of these, only the PMd and pre-sma contralateral to the working hand constantly showed preferential activation during Task 1 in comparison with Task 2 in all the subjects (Table 1). Table 1 also shows that, in the PMd and pre-sma, there were fewer pixels activated when Task 1 and Task 2 were compared than when Task 1 and Rest were compared. Since the two tasks were performed in a fixed order, the data could reflect nonspecific time effects. We examined this possibility by comparing the Rest conditions following Task 1 with that following Task 2 (see METHODS). However, no significantly activated pixels at P 0.005 (corrected for multiple comparisons) were found in any part of the obtained images, including the PMd, pre-sma, SMA, and MI. Figure 2 shows that the signal levels of the ROI in the PMd were similar during the two Rest conditions. DISCUSSION Consistent with previous activation studies (Fink et al. 1997; Grafton et al. 1998; Hazeltine et al. 1997; Iacoboni et al. 1998),

CORTICAL ACTIVATION DURING A CONDITIONAL MOTOR TASK 1669 FIG. 1. Activation maps from a representative subject performing the tasks. The numbers along the 5 serial slices indicate their z levels above the AC-PC line. The activated pixels with a threshold at P 0.005 (corrected for multiple comparisons) are color-coded according to the color bar on the right. The arrowheads in the two horizontal images (49.5 and 46.2) in the third column indicate the location of the VCA line. The sagittal images at the bottom show the location of the activated pixels with reference to the VCA line. The short lines around the skull on the sagittal images indicate the z levels of the 5 horizontal slices recorded in this study. AC, anterior commissure; Cent, central sulcus; L, left hemisphere; PC, posterior commissure; R, right hemisphere.

1670 K. KURATA, T. TSUJI, S. NARAKI, M. SEINO, AND Y. ABE TABLE 1. Location of activated areas in the frontal cortex showing statistically significant activity and number of pixels in each area Areas of Activation Coordinates (x, y, z) Task 1-Task 2 Task 1-Rest Task 2-Rest L lateral frontal BA4 MI 35.5 10.2 52.5 NS 7.94 (360) 8.08 (341) L lateral frontal BA6 PMd 31.8 6.1 54.2 6.66 (71) 8.60 (401) 8.42 (380) L medial frontal BA6 SMA 2.3 9.5 53.6 NS 7.80 (187) 7.74 (172) L medial frontal BA6 Pre-SMA 5.6 3.8 52.2 3.55 (21) 7.47 (247) 6.23 (212) All the numbers are average of five subjects. The coordinates are the center of mass of the entire activated area. Because there was no spatial normalization of the data, all the coordinates are subject-dependent. The numbers before parentheses indicate Z score greater than 3.09 in the areas in at least four of five subjects. The numbers in parentheses indicate number of significantly activated pixels in each area. L, left; BA, Brodmann s area; MI, primary motor cortex; PMd, dorsal premotor cortex; SMA, supplementary motor area; NS, not significant. we found that multiple regions in the frontal cerebral cortex, including the MI, PMd, SMA, and pre-sma, were activated during the two-finger movement tasks. The main, and new, finding of this study is that, of the cortical frontal motor areas activated, only the PMd and pre-sma contralateral to the working right hand were preferentially activated when auditory conditional cues were presented to select a required movement. The PMd and pre-sma were more active in the conditional task (Task 1) than in the control task (Task 2). The MI and SMA were more active during the two tasks than during Rest, but they failed to show statistically significant difference in activity between the two tasks. The overall number of auditory stimuli and movements executed in the two tasks over the entire session matched. When combined with previous evidence that the PMd and pre-sma are activated when arbitrary or spatial visual cues or spatial auditory cues are presented for motor selection (Fink et al. 1997; Grafton et al. 1998; Hazeltine et al. 1997; Iacoboni et al. 1998), and also activated when learning of visuomotor association is in progress (Deiber et al. 1997; Hazeltine et al. 1997; Sakai et al. 1999), our results strongly suggest that the PMd and pre-sma play a more extensive, general role in sensorimotor integration than has previously been thought. It is likely that the areas will also show preferential activation when conditional cues with any other modalities, such as somatosensory cues, are transformed into the appropriate motor behavior. Our observations also imply that the PMd and pre-sma contain highly modifiable neural circuits, known as hidden units (Churchland and Sejnowski 1992; Fetz 1992) that can switch connections between various inputs and outputs for conditional motor behavior. FIG. 2. Time course of the mean signal intensity of the region of interest (ROI) selected in the left dorsal premotor cortex (PMd) during the entire scan of the subject in Fig. 1 while performing the two tasks. The ROI for the PMd is indicated by a red square in the left panel (Task 1 versus Task 2, z 56.1 mm; see also Fig. 1).

CORTICAL ACTIVATION DURING A CONDITIONAL MOTOR TASK 1671 It has been reported that a number of neurons in the PMd of monkeys showed sustained change after the presentation of visual and auditory instruction cues, termed set-related activity (Boussaoud and Wise 1993; di Pellegrino and Wise 1993; Kurata 1993; Kurata and Hoffman 1994; Kurata and Wise 1988; Weinrich and Wise 1982). When arbitrary versus spatial visual cues indicated the same movements, a number of setrelated neurons showed higher modulation of activity following the conditional cues than after the spatial cues (Kurata and Hoffman 1994; Kurata and Wise 1988). Furthermore, when muscimol was injected to inactivate the area of the PMd where the set-related neurons were densely located, the monkeys showed severe deficits in selecting correct movement when conditional, but not spatial, cues were presented (Kurata and Hoffman 1994). Thus, it has been suggested that the PMd contributes not only to preparation for the movements in the conditional behavior, but also to selection of the movements (Kurata and Hoffman 1994; Kurata and Wise 1988). This view is further supported by the preferential activation of the human PMd during the conditional task used in this study. In this study, the activated area of the PMd was rostrally adjacent to the MI, a region termed the caudal PMd (PMdc) (Grafton et al. 1998; Iacoboni et al. 1998; Matelli et al. 1985). It was in the PMdc, not the rostral PMd (PMdr), that preferential changes in neuronal activity were observed in monkeys, and inactivation resulted in behavioral deficits during conditional tasks. This is also where activation during movement selection based on visual conditional cues was found in humans (Grafton et al. 1998). Although the PMdr is reportedly activated when spatial auditory cues are presented for motor responses (Iacoboni et al. 1998), we did not confirm this in our study, perhaps due to the conditional nature of the auditory cues. In monkeys, the PMdr seems to be more involved in oculomotor control than in limb movements (Fujii et al. 2000; Mitz and Godschalk 1989). We found that the pre-sma was more activated during the conditional task than during the control task, in which the subjects were required to select their finger movements internally. These results contrast with a report that the pre-sma is activated with self-initiated rather than with visually triggered finger movements (Deiber et al. 1999). However, the difference in the results can be explained by the design of the tasks. Deiber et al. (1999) used a visual cue as the only trigger without any conditional association, whereas we used conditional auditory cues, not visual ones. The pre-sma of monkeys contains abundant neurons with phasic responses to visual cue signals indicating the direction of forthcoming movements (Matsuzaka et al. 1992), and the human pre-sma is suggested to play a role in spatial visuomotor association (Sakai et al. 1999), which supports our findings. In interpreting our data, preferential cortical activation resulting from behavioral effects other than the conditional aspects of the tasks should be considered. For example, visuospatial attention modulates activity of the frontal motor areas in humans (Corbetta et al. 1993; Hazeltine et al. 1997; Nobre et al. 1997) and in monkeys (Boussaoud and Wise 1993; di Pellegrino and Wise 1993). Although the location of the areas modulated by visuospatial attention seems different in previous studies from the areas activated in this study, whether the PMd and other motor areas are modulated by auditory attention should be examined directly in the future. The subjects in our study might have been less attentive during the control task than during the conditional task, or inhibition of conditional behavior could have occurred during the control task. However, it was confirmed that the motor areas were activated in the two tasks, and preferential activation was consistently observed in the PMd and pre-sma, but not in the MI and SMA. Thus, it is more likely that specific changes do occur in the PMd and pre-sma during the conditional task. This work was supported by grants from the Ministry of Education, Science, and Culture of Japan (09268204, 10164205, and 11145203), Research for the Future Program (96L00206) by the Japan Society for the Promotion of Science, and Karouji Foundation. REFERENCES BOUSSAOUD D AND WISE SP. Primate frontal cortex: neuronal activity following attentional versus intentional cues. Exp Brain Res 95: 15 27, 1993. CHURCHLAND PS AND SEJNOWSKI TJ. The Computational Brain. Cambridge, MA: The MIT Press, 1992. CORBETTA M, MIEZIN FM, SHULMAN GL, AND PETERSON SE. A PET study of visuospatial attention. J Neurosci 13: 1202 1226, 1993. DEIBER M-P, HONDA M, IBANEZ V, SADATO N, AND HALLETT M. Mesial motor areas in self-initiated versus externally triggered movements examined with fmri: effect of movement type and rate. J Neurophysiol 81: 3065 3077, 1999. DEIBER MP, WISE SP, HONDA M, CATALAN MJ, GRAFMAN J, AND HALLETT M. Frontal and parietal networks for conditional motor learning: a positron emission tomography study. J Neurophysiol 78: 977 991, 1997. DI PELLEGRINO G AND WISE SP. Visuospatial versus visuomotor activity in the premotor and prefrontal cortex of a primate. J Neurosci 13: 1227 1243, 1993. FETZ EE. Are movement parameters recognizably coded in activity of single neurons? Behav Brain Sci 15: 679 690, 1992. FINK GR, FRACKOWIAK RS, PIETRZYK U, AND PASSINGHAM RE. Multiple nonprimary motor areas in the human cortex. J Neurophysiol 77: 2164 2174, 1997. FREUND H-J. What is the evidence for multiple motor areas in the human brain. In: Motor Control: Concepts and Issues, edited by Humphrey DR and Freund H-J. Chichester, UK: Wiley, 1991, p. 399 411. FUJII N, MUSHIAKE H, AND TANJI J. Rostrocaudal distinction of the dorsal premotor area based on oculomotor involvement. J Neurophysiol 83: 1764 1769, 2000. GRAFTON ST, FAGG AH, AND ARBIB MA. Dorsal premotor cortex and conditional movement selection: a PET functional mapping study. J Neurophysiol 79: 1092 1097, 1998. HALSBAND U AND FREUND H-J. Premotor cortex and conditional motor learning in man. Brain 113: 207 222, 1990. HAZELTINE E, GRAFTON ST, AND IVRY R. Attention and stimulus characteristics determine the locus of motor sequence encoding: a PET study. Brain 120: 123 140, 1997. HIKOSAKA O, SAKAI K, MIYAUCHI S, TAKINO R, SASAKI Y, AND PÜTZ B. Activation of human presupplementary motor area in learning of sequential procedures: a functional MRI study. J Neurophysiol 76: 617 621, 1996. IACOBONI M, WOODS RP, AND MAZZIOTTA JC. Bimodal (auditory and visual) left frontoparietal circuitry for sensorimotor integration and sensorimotor learning. Brain 121: 2135 2143, 1998. KURATA K. Premotor cortex of monkeys: set- and movement-related activity reflecting amplitude and direction of wrist movements. J Neurophysiol 69: 187 200, 1993. KURATA K AND HOFFMAN DS. Differential effects of muscimol microinjection into dorsal and ventral aspects of the premotor cortex of monkeys. J Neurophysiol 71: 1151 1164, 1994. KURATA K AND HOSHI E. Reacquisition deficits in prism adaptation after muscimol microinjection into the ventral premotor cortex of monkeys. J Neurophysiol 81: 1927 1938, 1999. KURATA K AND WISE SP. Premotor cortex of rhesus monkeys: set-related activity during two conditional motor tasks. Exp Brain Res 69: 327 343, 1988. MATELLI M, LUPPINO G, AND RIZZOLATTI G. Patterns of cytochrome oxidase activity in the frontal agranular cortex of the macaque monkey. Behav Brain Res 18: 126 136, 1985.

1672 K. KURATA, T. TSUJI, S. NARAKI, M. SEINO, AND Y. ABE MATSUZAKA Y, AIZAWA H, AND TANJI J. A motor area rostral to the supplementary motor area (presupplementary motor area) in the monkey: neuronal activity during a learned motor task. J Neurophysiol 68: 653 662, 1992. MITZ AR AND GODSCHALK M. Eye-movement representation in the frontal lobe of rhesus monkeys. Neurosci Lett 106: 157 162, 1989. MITZ AR, GODSCHALK M, AND WISE SP. Learning-dependent neuronal activity in the premotor cortex: activity during the acquisition of conditional motor associations. J Neurosci 11: 1855 1872, 1991. NOBRE AC, SEBESTYEN GN, GITELMAN DR, MESULAM MM, FRACKOWIAK RS, AND FRITH CD. Functional localization of the system for visuospatial attention using positron emission tomography. Brain 120: 515 533, 1997. PASSINGHAM RE. Cortical mechanisms and cues for action. Philos Trans R Soc Lond B Biol Sci 308: 101 111, 1985a. PASSINGHAM RE. Premotor cortex: sensory cues and movement. Behav Brain Res 18: 175 185, 1985b. PETRIDES M. Conditional learning and the primate frontal cortex. In: The Frontal Lobe Revisited, edited by Perecman E. New York: The IRBN Press, 1987, p. 91 108. SAKAI K, HIKOSAKA O, MIYAUCHI S, SASAKI Y, FUJIMAKI N, AND PÜTZ B. Presupplementary motor area activation during sequence learning reflects visuomotor association. J Neurosci 19: RC1, 1999. TALAIRACH J AND TOURNOUX P. Co-Planar Stereotaxic Atlas of the Human Brain. New York: Thieme Medical, 1988. VOROBIEV V, GOVONI P, RIZZOLATTI G, MATELLI M, AND LUPPINO G. Parcellation of human mesial area 6: cytoarchitectonic evidence for three separate areas. Eur J Neurosci 10: 2199 2203, 1998. WEINRICH M AND WISE SP. The premotor cortex of the monkey. J Neurosci 2: 1329 1345, 1982. WISE SP, ALEXANDER GE, ALTMAN JS, BROOKS VB, FREUND H-J, FROMM CJ, HUMPHREY DR, SASAKI K, STRICK PL, TANJI J, VOGEL S, AND WIESENDAN- GER M. Group report: what are the specific functions of the different motor areas? In: Motor Control: Concepts and Issues, edited by Humphrey DR and Freund H-J. Chichester, UK: Wiley, 1991, p. 463 485. WISE SP, DI PELLEGRINO G, AND BOUSSAOUD D. The premotor cortex and nonstandard sensorimotor mapping. Can J Physiol Pharmacol 74: 469 482, 1996.