The Journal of Neurophysiology Vol. 79 No. 2 February 1998, pp. 1070-1080
Copyright ©1998 by the American Physiological Society

Role of the Human Rostral Supplementary Motor Area and the Basal Ganglia in Motor Sequence Control: Investigations With H₂ ¹⁵O PET

H. Boecker^{1, 2}, A. Dagher¹, A. O. Ceballos-Baumann^{1, 2}, R. E. Passingham³, M. Samuel¹, K. J. Friston¹, J.-B. Poline¹, C. Dettmers¹, B. Conrad², and D. J. Brooks¹

¹ Medical Research Council Cyclotron Unit, Hammersmith Hospitals, London W12 OHS; ² Department of Neurology, Technical University of Munich, D-81675 Munich, Germany; and ³ Department of Experimental Psychology, University of Oxford, Oxford OX1 3UD, United Kingdom

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Boecker, H., A. Dagher, A. O. Ceballos-Baumann, R. E. Passingham, M. Samuel, K. J. Friston, J.-B. Poline, C. Dettmers, B. Conrad, and D. J. Brooks. Role of the human rostral supplementary motor area and the basal ganglia in motor sequence control: investigations with H₂ ¹⁵O PET. J. Neurophysiol. 79: 1070-1080, 1998. The aim of this study was to investigate the functional anatomy of distributed cortical and subcortical motor areas in the human brain that participate in the central control of overlearned complex sequential unimanual finger movements. On the basis of previous research in nonhuman primates, a principal involvement of basal ganglia (medial premotor loops) was predicted for central control of finger sequences performed automatically. In pertinent areas, a correlation of activation levels with the complexity of a motor sequence was hypothesized. H₂ ¹⁵O positron emission tomography (PET) was used in a group of seven healthy male volunteers [mean age 32.0 ± 10.4 yr] to determine brain regions where levels of regional cerebral blood flow (rCBF) correlated with graded complexity levels of five different key-press sequences. All sequences were overlearned before PET and involved key-presses of fingers II-V of the right hand. Movements of individual fingers were kept constant throughout all five conditions by external pacing at 1-Hz intervals. Positive correlations of rCBF with increasing sequence complexity were identified in the contralateral rostral supplementary motor area (pre-SMA) and the associated pallido-thalamic loop, as well as in right parietal area 7 and ipsilateral primary motor cortex (M1). In contrast, while rCBF in contralateral M1 and and extensive parts of caudal SMA was increased compared with rest during task performance, significant correlated increases of rCBF with sequence complexity were not observed. Inverse correlations of rCBF with increasing sequence complexity were identified in mesial prefrontal-, medial temporal-, and anterior cingulate areas. The findings provide further evidence in humans supporting the notion of a segregation of SMA into functionally distinct subcomponents: although pre-SMA was differentially activated depending on the complexity of a sequence of learned finger movements, such modulation was not detectable in caudal SMA (except the most antero-superior part), implicating a motor executive role. Our observations of complexity-correlated rCBF increases in anterior globus palllidus suggest a specific role for the basal ganglia in the process of sequence facilitation and control. They may act to filter and focus input from motor cortical areas as patterns of action become increasingly complex.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Complex sequential motor actions such as piano playing or typewriting involve a sophisticated coordination of multiple predetermined movements according to a specific temporo-spatial context. There is evidence from single unit recordings in nonhuman primates that the supplementary motor area (SMA), in particular its rostral part (pre-SMA), subserves the central control of internally generated sequential limb movements (Halsband et al. 1994; Tanji 1994; Tanji and Shima 1994). Moreover, data derived from investigations of movement-related pallidal discharges in monkeys indicate that the basal ganglia may act to focus desired motor patterns by inhibiting unwanted movements (Marsden and Obeso 1994; Mink and Thach 1991) and to facilitate sequential motor actions by sending signals to the SMA after each step of a predictable movement sequence (Brotchie et al. 1991; Marsden and Obeso 1994). The temporal characteristics of pallidal discharges (Brotchie et al. 1991) in relation to subcomponents of movement sequences are consistent with the notion that "the motor regions of the basal ganglia deliver instructions, based upon a read-out of ongoing activity in sensorimotor cortex, to premotor areas in such a way as to set up the correct motor programmes required for the next motor actions" (Marsden 1987).

An approach toward understanding the functional anatomy of higher order motor control mechanisms in the human brain is to investigate the localization of changes in regional cerebral blood flow (rCBF), an index of local synaptic activity, associated with movements requiring different levels of central planning and control. Previous functional imaging studies have utilized a variety of motor paradigms, ranging from simple ballistic to highly complex sequences of finger movement (Boecker et al. 1994; Colebatch et al. 1991; Fox et al. 1985; Matelli et al. 1993; Rao et al. 1993; Remy et al. 1994; Roland et al. 1980a, 1982; Sergent et al. 1992; Shibasaki et al. 1993; Tyszka et al. 1994). These studies have supported the notion that the SMA is involved in the process of organizing sequential motor actions, as have studies of remembered and voluntary saccade generation (Anderson et al. 1994; O'Sullivan et al. 1995; Petit et al. 1993) and speech (Roland 1984).

The concept of a functional segregation of human SMA into distinct subcomponents, with a rostral area particularly involved in higher order motor planning and control, is supported by recent neuro-imaging studies. Deiber et al. (Deiber et al. 1991) and Playford et al. (Playford et al. 1992) demonstrated selective activation of rostral SMA (i.e., region anterior to the cortical projection of the anterior commissure) along with dorsolateral prefrontal cortex (DLPFC) when subjects performed joystick movements in freely selected directions, paced by a tone, compared with cued or stereotyped movements in a fixed forward manner. More recent imaging studies with magnetic resonance imaging (MRI) and positron emission tomography (PET) indicate that pure motor imagination, as opposed to motor performance, is associated with enhanced rostral activation of the mesial premotor cortex (Stephan et al. 1995; Tyszka et al. 1994). These data indicate a rostro-caudal segregation of premotor areas involved in motor planning as opposed to motor execution. Nevertheless structure-function relationships in human premotor cortical areas remain incompletely understood, in particular, the functional interactions between basal ganglia and cortical motor areas in the process of motor sequence control.

A way of approaching these issues more precisely with PET is to define characteristics of task-dependent central processing via correlational analyses that relate evoked rCBF responses to graded modulations of specific performance parameters (Frackowiak and Friston 1994; Friston et al. 1993). Previous correlational analyses with modulations of basic movement parameters have demonstrated that graded increases in frequency and force levels of movement are associated with increases of rCBF in primary motor cortex (M1) and caudal SMA, but not in pre-SMA nor in the basal ganglia (Blinkenberg et al. 1996; Dettmers et al. 1995; Jenkins et al. 1997; Sadato et al. 1996b). These findings support the notion that caudal SMA has a motor executive role closely linked to M1, in contrast to rostral SMA and basal ganglia that do not seem to be involved in determining basic parameters of movement execution.

On the basis of this recent work, we wished to identify those parts of the distributed motor system that control and facilitate complex patterns of sequential movement as opposed to those parts that are primarily executive. A recently published study has addressed this issue by correlating rCBF with units of sequence length as an index of "complexity," whereas rate, rhythm, and total number of movements were kept constant (Sadato et al. 1996a). Sequence length increased linearly from 4 (complexity level I) to a maximum of 16 (complexity level IV) opponent finger movements, performed on the basis of memory. In this study, positive correlations with sequence length were identified in right dorsal premotor cortex (Brodmann area 6) and right precuneus (Brodmann area 7). Sadato and coworkers interpreted their findings as an indication that these areas play a role in the "storage of motor sequences in spatial working memory and the production of ongoing sequential movement with reference to that of buffered memory."

We use a similar methodological approach, but rather than primarily interrogating mechanisms of retrieval from motor memory, our PET study focuses on the central control of motor sequences that increase with respect to the "complexity" of actual motor performance. According to our study design, this is accomplished by introducing repetitive and interposed finger movements that render consecutive movement sequences more complex compared with a simple unidirectional sequence of four finger movements in order. Given the difficulty in performing these motor tasks accurately, all five motor sequences were overlearned before PET, to reduce any effect of motor learning in the course of this study.

On the basis of the electrophysiological recording data in monkeys performing prelearned sequences of hand movements (see previous text), we hypothesized primarily rostral SMA (pre-SMA) and basal ganglia to be involved in the central control of automated sequential movement. Additional involvement of the ipsilateral primary motor cortex was anticipated on the basis of recent imaging data comparing sequential and ballistic motor tasks (Shibasaki et al. 1993).

	METHODS

Abstract Introduction Methods Results Discussion References

Subjects

We scanned seven healthy male volunteers with a mean age of 32.0 ± 10.4 yr in the study. None of them had a history of neurological or psychiatric disease and none were taking medication at the time of PET. All subjects were right-handed, as assessed by the Edinburgh Handedness Inventory (Oldfield 1971). Written informed consent was given by all subjects before the PET experiment. The studies were approved by the Hammersmith Hospital Medical Ethics Committee. Permission to administer radioactivity was obtained from the Administration of Radioactive Substances Advisory Committee (ARSAC) of the Department of Health of the United Kingdom.

Experimental design

Repeated measurements of rCBF were performed in each subject with H₂ ¹⁵O PET. Each 12-run PET study consisted of two resting conditions and five pairs of key-press sequences involving fingers II-V of the right hand. Scans were arranged according to a Latin-Square design to minimize potential time and order effects. Subjects had to continue performing each sequence repeatedly for the duration of each scan. Finger movements in all five sequences were kept constant by external pacing with a computer-generated sound at one second intervals but sequence lengths ranged from four to a total of eight finger movements of increasing complexity (see Table 1). Sequences differed in terms of the number of repetitive and interposed finger movements in the opposite direction added to a simple unidirectional sequence of all four fingers in order, corresponding to sequence one. To avoid internal calculation as a confounding factor during the PET experiment, detailed verbal descriptions of the required movements were given for each of the five sequences. In particular, subjects were asked not to assign numbers to finger movements

To remove motor learning during the course of this study, all subjects recruited for the PET scan were instructed two weeks before the experiment to regularly rehearse all five sequences of finger movements until performance was so automatic that they were able to read a book or maintain a conversation while performing each of the tasks. On the day of PET, the degree of automaticity was tested with serial object span tasks according to a previously published report (Jenkins et al. 1994): while performing each of the five sequence tasks separately, five different independent objects were named at a rate of one per second and subjects required to repeat them instantaneously in correct order. Adequate performance in this test was a prerequisite for subsequent PET scanning.

During PET, response times of individual finger movements were recorded throughout all five motor tasks. Subjects were explicitly instructed to avoid anticipation, i.e., not to move their finger before the onset of the next pacing tone in the sequence. On-line video recording was used to monitor any associated movements of the left hand. In addition, separate surface electromyographical recordings were performed in two volunteers under identical experimental conditions to investigate task-related coactivation of the ipsilateral proximal and distal arm musculature.

Data acquisition

Subjects were scanned while lying supine with their eyes closed and their ears unplugged. Head fixation was accomplished with individually fitted vacuum-operated polysterene head moulds. Head positioning was carried out by aligning the transaxial PET planes to the orbitomeatal line with a laser beam. Head position was then adjusted carefully to include the vertex in the upper limit of the data set. Therefore inclusion of the SMA in the field of view (FOV) of the scanner was ensured in all subjects. Because the FOV of our camera is 10.65 cm, we were not able to include the entire cerebellum in this study and so cerebellar activity will not be addressed further in this manuscript.

Measurements of regional distribution of radioactivity were performed with a CTI 953 B PET camera (CTI, Knoxville, TN). The physical characteristics of the camera have been described previously (Spinks et al. 1992). Interplane septa were retracted to acquire data in the three-dimensional mode and improve point source sensitivity (Bailey et al. 1991). After corrections for randoms, dead time, and scatter, all emission data were reconstructed by filtered back projection (Hanning filter, 0.5 cycles/pixel cutoff frequency) to 31 consecutive axial planes with a resolution of 8.5 × 8.5 × 4.3 mm at full-width at half-maximum (FWHM) (Spinks et al. 1992). Reconstructed images were displayed in a matrix consisting of 128 × 128 × 31 voxels with individual voxel sizes measuring 2.09 × 2.09 × 3.43 mm.

Before collecting activation data, 20 min ⁶⁸Ga/⁶⁸Ge-derived transmission scans were recorded individually and used to correct for attenuation of the 511 KeV radiation by skull and brain. Then the distribution of cerebral radioactivity during each scan was recorded over 90 s; each measurement was started with a background scan of 30 s. After a delay of about 30 s, depending on individual circulation times, 11.5 mCi/run of freely diffusible positron-emitting water (H₂ ¹⁵O) in 3 ml of normal saline solution were administered by an automated infusion pump as a slow bolus (20 s) into the subject's left antecubital vein (Silbersweig et al. 1993). Scanning started ~5 s before the onset of rise of the background counts, ~30 s after the start of infusion. Each experimental condition was started ~10 s before data acquisition and was continued until the completion of the scan. This process was repeated for all of the 12 scans, with between-scan intervals of 10 min to allow time for radioactive decay.

Data transformation

All calculations and image transformations were performed on Sun SPARC 5 workstations (Sun Computers Europe, Surrey, UK). For data analysis, statistical parametric mapping software (SPM 95, Wellcome Department of Cognitive Neurology, London) implemented in the Matlab environment (Mathworks, Sherborn, MA) was used. SPMs are spatially extended statistical processes that are used to characterize regionally specific effects in imaging data. SPM combines the general linear model and the theory of Gaussian fields to make statistical inferences about regional effects (Friston et al. 1991, 1994; Worsley et al. 1992). In the first stage of analysis, the scans of each subject were coaligned to the first of the series by using an automated realignment program based on a six parameter rigid-body transformation using a least-squares technique on a voxel-by-voxel basis (Friston et al. 1995a). This generated an aligned set of images per subject and a mean image (each of 31 planes). The mean image, characterized by the highest anatomic detail, was transformed into standard stereotactic space, corresponding to the atlas of Talairach and Tournoux (Talairach and Tournoux 1988). This procedure of spatial normalization involves a 12 parameter affine (linear) and quadratic (nonlinear) three-dimensional transformation on a slice-by-slice basis (Friston et al. 1995a). The same transformations were then applied to each of the subject's realigned images to ensure identical orientation in standard space. As a final preprocessing step, all data were smoothed with an isotropic Gaussian kernel of FWHM 12 mm to increase the signal-to-noise ratio and to compensate for individual differences in gyral anatomy.

Data analysis

For data analysis, the condition, subject, and covariate effects were estimated according to the general model at each voxel. The confounding effect of variation in global blood flow across subjects and scans was removed by an analysis of covariance (ANCOVA) with global flow as the confounding variable and mean rCBF normalized to an arbitrary level of 50 ml·100 ml¹·min¹ (Friston et al. 1995b). These adjusted rCBF voxel values were then used for further statistical analyses.

The statistical analyses of the group data in this study comprised formal categorical comparisons and independent correlational analyses. To identify the exact location of brain areas involved in the task per se, categorical comparisons comparing the average of all activation scans with the resting state were performed (P < 0.05, corrected for multiple nonindependent observations). Each of these activated and deactivated brain areas were then interrogated to determine whether or not rCBF correlated with increasing sequence complexity. For this second analysis, the resting state scans were not included. The X, Y, and Z coordinates of activation foci identified by the categorical comparisons were used to test for correlated rCBF changes. A maximum deviation of 1 cm was allowed to identify voxels with peak correlation within the same anatomic region. Because the areas investigated had already been identified as participating in the task (i.e., to belong to the distributed motor circuit that is activated or deactivated during performance of complex sequential movements) a lower threshold for significance was set (P < 0.01) for correlated rCBF changes and no correction for multiple comparisons was applied.

This dual approach has the potential to functionally discriminate primarily executive motor regions from hierarchically higher motor areas that are specifically concerned with the central control of complex sequential movement. Because basic motor parameters (e.g., frequency, amplitude of movement) were controlled by the task design, we expected essentially executive motor regions to behave monotonically throughout all five conditions, in contrast to central higher motor control areas where modulated rCBF increases relative to sequence complexity were predicted.

	RESULTS

Abstract Introduction Methods Results Discussion References

The mean response times of individual finger movements, recorded for the five sequential finger movement tasks, were 0.35 ± 0.02 s. Differences in the mean response times among individual group means for sequences I-V were nonsignificant [P = 0.49, repeated measures analysis of variance(ANOVA)]. Moreover, no differences in task error rates were recorded. On-line video recording revealed no visible movements of the left hand in any of the subjects participating in the study. Furthermore, separate surface electromyographical recordings that were performed outside the PET scanner in two further volunteers under identical experimental conditions revealed no task-related coactivation of the ipsilateral arm musculature, either proximally or distally.

View larger version (71K): [in this window] [in a new window]   FIG. 1. Surface projections (A and C) and axial sections (B) of color-coded SPMs superimposed onto stereotactically normalized (Talairach and Tournoux 1988) T1-weighted magnetic resonance imaging (MRI) images. Left side of figure corresponds to right hemisphere. A: surface-rendering of statistical parametric maps (SPMs) showing categorical (left) and correlated (right) increases of regional cerebral blood flow (rCBF). At a threshold of P < 0.001, only left pre-supplementary motor area (pre-SMA) and right parietal area 7 show correlated increases of rCBF with sequence complexity. These regions constitute a subset of widely activated cortical premotor, primary motor, and association areas as identified by categorical comparison. B: axial cortical section (56 mm above bicommissural line) demonstrating precise anatomic location of peak categorical (top, P < 0.001) and correlated (bottom) rCBF increases in SMA with respect to vertical line passing through anterior commissure (red horizontal line). Threshold of bottom image was reduced to P < 0.05 to demonstrate both spatial extent of correlated pre-SMA focus of rCBF increase and absence of any such correlation in contralateral M1 and caudal posterior SMA. Voxels with highest correlations of rCBF and sequence complexity are located in contralateral pre-SMA, rostral to anterior commissure. C: surface-rendering of SPMs demonstrating categorical (left) and correlated (right) decreases of rCBF (P < 0.001). Highest correlated decreases of rCBF with increasing sequence complexity are located in right mesial prefrontal area 10 and right medial temporal area 39.

Areas significantly activated by the sequential finger task with the right hand compared with the resting condition included the contralateral (Z score, 9.23) and ipsilateral (Z score, 4.07) primary sensorimotor cortex, lateral, and medial premotor areas (covering both rostral and caudal SMA), anterior cingulate, parietal area 40, and precuneus. Subcortically, basal ganglia were activated bilaterally along with left thalamus (Table 2A). Areas significantly deactivated by the tasks (Table 3A) included secondary visual cortex, inferior- and medial temporal gyri bilaterally, and anterior and posterior cingulate, as well as frontal cortical areas (Brodmann areas 10 and 8).

Correlated increases of rCBF with sequence complexity (Table 2B) were seen in a subset of distributed motor areas identified by the categorical comparison (Fig. 1A,B; Fig. 2). These areas were the rostral SMA (pre-SMA, extending into the antero-superior part of caudal SMA), ipsilateral M1, right parietal area 7, left thalamus, and bilateral anterior globus pallidus, extending into adjacent putamen. In contrast, although contralateral M1 and extensive parts of caudal SMA showed significant activation with movement compared with rest, changes in rCBF did not correlate with task complexity (Fig. 1, A and B).

View larger version (54K): [in this window] [in a new window]   FIG. 2. SPMs of subcortical areas showing correlation of rCBF with sequence complexity. Group data are superimposed onto three consecutive (from left to right: 12, 8, and 4 mm above bicommissural line) stereotactically normalized T1-weighted MRI images (Talairach and Tournoux 1988). For display purposes, all results are thresholded at P < 0.05, uncorrected. Correlated changes in rCBF were identified bilaterally in anterior globus pallidus (peak correlation, extending into adjacent putamen) and in contralateral thalamus. Z-scores are represented by color-coded bar on right.

Correlated decreases of rCBF with sequence complexity were seen in subsets of brain areas deactivated compared with rest (Table 3B). Areas of significantly correlated rCBFdecreases included mesial prefrontal-, right medial temporal-,and anterior cingulate (Brodmann area 32) (Table 3B; Fig. 1C). Figure 3 shows scatter diagrams of adjusted individual rCBF levels in four regions of particular interest plotted against consecutive task complexity levels I-V.

View larger version (26K): [in this window] [in a new window]   FIG. 3. Scatter diagrams of normalized rCBF values in ml/100 g/min (ordinate) at rest (R) and vs. sequence complexity levels I-V (abcissa) are shown from voxels with peak correlation in (A) pre-SMA (X/Y/Z = 4/2/56), (B) contralateral sensorimotor hand area (X/Y/Z = 34/22/52), (C) right pallidum (X/Y/Z = 16/2/8), and (D) parieto-temporal area 39(X/Y/Z = 46/64/20). Each column contains seven pairs (i.e., number of subjects) of identically colored points that represent individual adjusted rCBF values. Anterior SMA shows progressive rCBF increases with increasing levels of sequence complexity, less significant increases are demonstrated in anterior globus pallidus. In contrast, contralateral primary sensorimotor cortex shows a constant level of increased rCBF with sequential movement compared with rest, independent of level of graded sequence complexity. Temporal area 39 shows correlated decreases of adjusted rCBF levels with sequence complexity.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

The findings of this study indicate that neuronal activity within a distributed cortical and subcortical motor system can be attributed to the overall complexity of a familiar motor task. The pattern of activated brain areas obtained on categorical analysis confirms previous imaging data on sequential finger movements (Rao et al. 1993; Roland 1985; Shibasaki et al. 1993; Wessel et al. 1995), including widespread bilateral activation of motor-, premotor-, and parietal cortical areas, along with basal ganglia and thalamus. Participation of cerebellar structures in motor sequencing is also to be expected on the basis of previous PET studies (Fox et al. 1985; Shibasaki et al. 1993; Wessel et al. 1995) but these structures were not sufficiently covered by the field of view of our PET camera.

Of the cortical and subcortical areas of the brain activated by sequential finger movements, only selected regions showed significant correlation of evoked rCBF responses with changes in sequence complexity. These brain areas can, therefore, be postulated to have a primary role in the central control and facilitation rather than in the execution of individual finger movements during sequences. In the following sections, the results of this study will be discussed area by area with respect to their involvement in sequence control.

Primary sensorimotor cortex

The noncorrelated increase of rCBF in the contralateral primary motor cortex during performance of increasingly complex sequential finger movement supports a role for M1 predominantly related to execution of individual finger movement during a sequence. Single-cell activity recordings from M1 in awake monkeys trained to perform a sequential motor task revealed very similar levels of neuronal firing, regardless of whether the motor task was visually guided or internally cued (Mushiake et al. 1991). Moreover M1 activity was similar when movements were determined by visual or auditory signals (Tanji and Kurata 1982). This lack of "modality specificity" (Tanji and Shima 1996) is one of the major features differentiating M1 from premotor cortical function (see below). Functional imaging data in humans support these conclusions. Jahanashani and coworkers, for example, recently reported similar levels of M1 activation during self-initiated and externally triggered movements (Jahanshahi et al. 1995). In line with our findings, Sadato and coworkers reported a lack of correlated activation levels in M1 as sequences of movements became increasingly complex. In their study, as in ours, there was careful control of basic movement parameters, in particular movement frequency (Sadato et al. 1996a).

Interestingly, we found complexity-correlated rCBF increases in ipsilateral M1, though at a weak statistical level of significance (Z score, 2.5; see Table 2B). Nevertheless, our findings support previous imaging data (Kim et al. 1993; Rao et al. 1993; Shibasaki et al. 1993), which have demonstrated involvement of ipsilateral M1 during complex sequential but not ballistic motor tasks. These observations, together with the identification of slow negative potential shifts over M1 before movement onset (Kitamura et al. 1993) as well as the demonstration of disturbed motor sequence performance after repetitive transcranial magnetic stimulation over the ipsilateral primary motor cortex (Chen et al. 1997), support the idea that ipsilateral M1 may be recruited for the control and preparation of complex sequential motor tasks. The absence of associated minor movements of the ipsilateral hand on video monitoring and electromyograph (EMG) suggests that correlated ipsilateral M1 activation was not artifactual.

Premotor cortex

The results of our study, both on categorical and correlational analyses, provide further in vivo evidence supporting a role of the mesial premotor cortex in the central processing of sequential limb movements, particularly those that are internally guided and performed on the basis of memory (Goldberg 1985; Halsband et al. 1993; Passingham 1987; Tanji 1994; Tanji and Shima 1994, 1996). This is in line with analyses of single cell SMA activity in trained monkeys (Mushiake et al. 1990) that demonstrate preferential activity of mesial premotor cortex during remembered (as compared with externally driven) sequences of movement. Recently, these recording data were further extended by the demonstration of "sequence-specific neurons" in SMA that are active only in response to one particular movement sequence and not an other (Mushiake et al. 1991; Tanji 1994; Tanji and Shima 1994, 1996; see below).

Electrophysiological evidence in humans supporting a role for SMA in the organization of sequential movement stems from readiness potential studies (Deecke et al. 1969), which have demonstrated highest early negative amplitudes over the vertex when finger movements are sequential or bimanual (Benecke et al. 1985; Kitamura et al. 1993; Lang et al. 1988). Similar conclusions can be drawn from pioneering functional imaging work with Xenon-SPECT, which demonstrated SMA activation during the execution of a sequence of different isolated finger movements, but not during fast repetitive (nonsequential) flexions of the same fingers (Roland et al. 1980a). More recent imaging studies with PET, however, using intersubject averaging, have shown participation of caudal SMA also during nonsequential motor tasks (Colebatch et al. 1991; Fox et al. 1985), findings that are further supported by recordings of movement-related potentials from subdural electrodes in patients performing nonsequential movements (Ikeda et al. 1992). Nevertheless, a variety of more recent imaging studies have unambiguously shown enhanced SMA activity during sequential or volitional as compared with simple cued ballistic (nonindividuated) tasks (Rao et al. 1993; Remy et al. 1994; Shibasaki et al. 1993). These data, together with the demonstration of SMA but not M1 activation during pure motor ideation (Roland et al. 1980b; Stephan et al. 1995) and evidence from SMA lesion studies (Freund 1987) are compatible with the human SMA playing a central role in the higher control of motor subroutines and the organization of motor actions in sequential order (Marsden et al. 1996; Tanji 1994; Tanji and Shima 1996).

A more refined concept of the functional organization of mesial premotor cortex has emerged from neuronal recordings in awake monkeys. These show increased single-unit activity in rostral SMA (i.e., pre-SMA) during preparatory periods before performance of prelearned movement (Matsuzaka et al. 1992; Tanji 1994). In contrast, direct movement-related activity, time-locked with the EMG signal, is more prominent in caudal SMA (Matsuzaka et al. 1992). Further work by the same group (Halsband et al. 1994) has extended these findings by studying predetermined sequences of movement either on the basis of retrieval (internally guided) or cued by external triggers (visually guided). Pre-SMA neurons are more active during movement preparation compared with execution. During movement they fire preferentially when the task is internally guided. These studies in monkeys demonstrate a functional subdivision of rostral and caudal SMA, which is supported by cytoarchitectonic- and histochemical investigations (Luppino et al. 1993; Matelli et al. 1991; Rizzolatti et al. 1996; Zilles et al. 1995, 1996) and intracortical microstimulation findings (Luppino et al. 1991; Matsuzaka et al. 1992). These have revealed reciprocal connections between caudal SMA and M1. In contrast, while the pre-SMA projects to caudal SMA, it is not directly connected with M1. Instead, pre-SMA has connections with anterior cingulate cortex (area 24), the DLPFC, anterior lateral premotor cortices, and posterior parietal areas, including parietal area 7 (Luppino et al. 1993).

The concept of a subdivision of SMA into functionally distinct parts along the rostrocaudal axis is supported by recent functional imaging studies in humans (for a recent overview see: Picard and Strick 1996), complemented by the data presented here. It has been considered that a functional subdivision between caudal and rostral (pre-) SMA occurs at the level of the ventral anterior commissure (VAC) line (Passingham 1996; Picard and Strick 1996). Taking such a boundary, pure motor imagery, for instance, is associated with activation located directly posterior to the VAC line, corresponding to the most rostral part of caudal SMA (Stephan et al. 1995). The first demonstration of selective pre-SMA activation in humans was obtained from a paced joystick task comparing freely selected to stereotyped forward movements (Deiber et al. 1991). In that study, pre-SMA and DLPFC, but not caudal SMA were relatively more activated, reflecting the process of decision making during the first but not the second task (Deiber et al. 1991; Playford et al. 1992). These findings were recently extended by work on internally selected versus externally cued finger movements (Deiber et al. 1996). The data associated with conscious selection processes can be contrasted with studies on simple repetitive upper limb movements not requiring any decision making, where the focus of SMA activation was located posterior to the VAC line (Colebatch et al. 1991). Moreover, investigations of graded variations of basic motor execution parameters, such as frequency and force of movement, have resulted in correlated rCBF increases in caudal SMA and contralateral M1, but not in pre-SMA (Blinkenberg et al. 1996; Dettmers et al. 1995; Jenkins et al. 1997; Sadato et al. 1996b).

In the context of these functional imaging data, our identification of rCBF increases in the pre-SMA territory that correlate with the complexity of a motor sequence is a novel finding. It complements the correlational data on basic motor execution parameters (see above) in that a task-specific correlated rCBF increase can, on the basis of our data, be attributed to the mesial premotor cortex territory located rostral to the VAC line. Because basic motor parameters were carefully controlled by the study design, our data suggest a role for pre-SMA, other than determining basic parameters of movement execution. Rather, our data, are to be interpreted in line with electrophysiological recording data in monkeys showing enhanced pre-SMA activity during preparation and control of prelearned movement sequences (Halsband et al. 1994; Picard and Strick 1996). Such control mechanisms appear to be required for adequate performance, in particular as sequencing becomes more complex. With respect to the work of Deiber and coworkers (Deiber et al. 1991), our demonstration of involvement of pre-SMA, but not DLPFC in a prelearned motor task suggests that pre-SMA may act to facilitate complex sequential movements even when no conscious decision making or selection process is required. A recent PET study comparing sequential with repetitive finger oppositions also identified enhanced SMA activation in motor sequencing (Shibasaki et al. 1993). However these authors found no relative anterior spread of activation toward rostral SMA in their sequential task. This difference between our findings and those of Shibasaki et al. may in part be attributed to variations in the task design (self-paced task, rather than cued) and also the analytic method employed (categorical rather than correlational approach).

It is important to point out that the absence of correlated rCBF increases in the bulk of caudal SMA (except for the most antero-superior part) does not mean that caudal SMA has a role entirely directed toward executive aspects of motor behavior. In fact, there is evidence from neuronal recording studies that sequence-specific neurons exist in caudal SMA (Tanji and Shima 1994). These neurons fire differentially depending on the specific order of similar movements in a sequence. These observations, along with lesion data, suggest that caudal SMA has a greater involvement than M1 with respect to central motor sequence control; this may not have been detected with PET for technical reasons.

Parietal area 7

The evoked rCBF responses observed in right posterior parietal area 7 (precuneus) correlated highly with the complexity of our sequential motor task. It is of interest that direct connections exist between pre-SMA and parietal area 7 (Luppino et al. 1993). This area is thought of as an integrative system involved in the processing of spatial aspects of movement. Imagining joystick movements (Stephan et al. 1995), tracking a target in space (Grafton et al. 1992b), and volitional saccadic eye movements in darkness (Petit et al. 1996) are all associated with activation in this area. Our findings suggest that subconscious internal visualization of finger movements in space may be occurring as sequences become increasingly complex. During the learning phase before PET, subjects were asked not to attribute numbers to a particular finger to avoid silent articulation or counting.

Sadato and coworkers identified similar correlated rCBF increases in right precuneus when sequence complexity increased (Sadato et al. 1996a). In interpreting their data, they discussed the possible influence of rising levels of attention required to process increasingly complex sequences.

Basal ganglia

The finding of complexity-correlated rCBF increases in the globus pallidus during our sequential motor task is of particular interest. Anatomically, the basal ganglia are part of a frontal-basal ganglia-thalamo-frontal loop. The output nucleus of the basal ganglia is the globus pallidus internus (GPi), which projects to the ventral thalamic nuclei (Albin et al. 1989; Flaherty and Graybiel 1994).

Activation of the basal ganglia has been demonstrated with functional imaging in most motor tasks but their precise role in the central organization and control of movement remains unclear. Previous PET studies have identified similar levels of activation of the lentiform nucleus when joystick movements are either stereotyped or freely selected (Deiber et al. 1991; Playford et al. 1992). There were also no differences in lentiform rCBF when movements were internally generated compared with externally paced (Jahanshahi et al. 1995; Jenkins et al. 1994a). During the course of motor learning tasks, lentiform activation stayed at a similar level suggesting that the basal ganglia do not primarily mediate motor procedural learning (Grafton et al. 1992a; Jenkins et al. 1994a). Furthermore, no correlation of lentiform rCBF with increasing movement frequency (Blinkenberg et al. 1996; Jenkins et al. 1997; Sadato et al. 1996b) or force (Dettmers et al. 1995) was found, suggesting that the basal ganglia do not determine basic movement parameters. For further discussion of these PET findings see (Brooks 1995).

There are however, electrophysiological data supporting a primary role of the basal ganglia in the process of facilitating and optimising sequential motor performance, this function being mediated through interaction with premotor cortex via thalamocortical projections. Brotchie and coworkers, (Brotchie et al. 1991) have recorded from pallidal cells in monkeys trained to perform predictable sequences of wrist movements. The firing of a subpopulation of these pallidal cells was characterized by a dependence on the degree of automaticity and predictability of sequential movements. A separate small group of pallidal neurons was characterized by biphasic responses at onset and cessation of EMG activity. The temporal characteristics of these pallidal discharges were interpreted as providing cues to the SMA, allowing it to switch to the next movement in a programmed sequence. Muscimol, a -aminobutyric acid (GABA) agonist, when injected into pallidum leads to both, tonic and phasic cocontraction of wrist flexors and extensors during movement (Mink and Thach 1991). This finding suggests that the pallidum may act to focus and filter desired motor patterns during movement, optimizing them and inhibiting unwanted movements (Marsden and Obeso 1994; Mink and Thach 1991). It is tempting, to interpret our finding of correlated rCBF increases in globus pallidus with complexity of sequential movement as further evidence that the basal ganglia mediate sequential movement, possibly via a focusing and filtering role.

Decreases in cortical activity

Correlated rCBF decreases in cortical areas (prefrontal and temporoparietal regions) with sequence complexity are likely to reflect activity reductions in brain regions irrelevant to the task. Indeed, rCBF was reduced in regions that are not pivotal to the overlearned sequential motor task. Jenkins and coworkers (Jenkins et al. 1994) have shown that neither prefrontal nor temporo-parietal activation is required during performance of learned sequences of finger movements with eyes closed. Our findings suggest that correlated rCBF decreases in nonmotor areas occur in parallel with correlated activity increases in the motor system. We suggest, therefore, that this is the result of task-dependent focusing of neuronal activity, which is essential for facilitating activation of the relevant brain circuitry.

Experimental design

The experimental design of this PET study allowed us, first, by externally pacing movements of all four fingers in each of the five sequences, to keep the total number of movements (i.e., "basic executive" motor parameters) constant throughout the entire study; second, by systematically changing only one of the experimental parameters, i.e., the overall complexity of the tested sequences or "higher-order" motor parameter, to identify areas involved in control of sequential movement. This distinguishes our study from previously published reports on comparisons of "complex" sequential with "simple" repetitive finger oppositions that differ not only in task complexity but also in terms of motor performance (Rao et al. 1993; Roland et al. 1980a; Shibasaki et al. 1993).

One recently published study has addressed the issue of motor sequence control by applying a similar correlational approach to ours. Sadato and colleagues, (Sadato et al. 1996a), have correlated levels of rCBF with proportionally increasing units of sequence length as an index of "complexity," while rate, rhythm, and total number of movements were kept constant. This study differs from ours, however, in the "a priori" definition of the independent variable used for correlation, i.e., proportionally graded increases of sequence length (Sadato et al. 1996a) as opposed to graded changes in sequence complexity and length applied in our study. The advantage of the design used by Sadato and coworkers is that the independent variable (sequence length) increases in a strictly linear fashion. Although in our design, repetitive and retrograde finger movements were introduced in a controlled fashion from sequences I-V, our sequence complexity increases were nonlinear. Our subjects reported that more training trials were required to attain automatic performance as sequences became increasingly complex, though we recorded no behavioral data to quantitate this. We feel confident, however, that the automatic design of our study focuses primarily on the central control necessary for the accurate performance of increasingly complex motor sequences, rather than the functional anatomy of retrieval of stored sequence information.

In support of this viewpoint, Sadato and coworkers found no correlated rCBF changes in SMA as sequences of movements became increasingly lengthy. They concluded that "activation of the SMA is related to the execution of the sequence regardless of its length or complexity" (Sadato et al. 1996a). These findings may be explained by the fact that, although sequence length was the independent variable, the tasks under investigation did not differ in terms of complexity of performance, i.e., they required similar levels of sequencing of the four fingers.

Other issues of relevance to our findings are the degree of motor learning and attention to task performance. Complex sequential finger movements as studied in this PET experiment are difficult to execute accurately and require repeated learning trials before performance is automatic. The response time measurements of individual finger movements revealed no significant differences between sequence complexity levels I-V, implicating that the tasks were overtrained at the time of PET scanning. Moreover, the absence of prefrontal activation supports automaticity of task performance because prefrontal cortex activation was observed in a previous study from our group during novel sequence learning (Jenkins et al. 1994). Furthermore, in interpreting our results, we have no reason to suppose that attention to the tasks differed systematically depending on the complexity of the five sequences tested. Jueptner et al. (Jueptner et al. 1997), in a recent study, demonstrated that attention to the performance of a prelearned sequence activates dorsal prefrontal cortex along with cingulate area 32. Finally, our results can also not be explained by a time effect, because the order of different movement sequences was randomized for every subject and throughout the entire study.

	ACKNOWLEDGEMENTS

  The authors thank Dr. N. Kahlil from the Dept. of Neurophysiology, Hammersmith Hospital, London for the electromyographic recordings. Furthermore, we thank A. Blyth, A. Williams, D. Griffith, and G. Lewington for expert technical assistance at the PET camera.

  H. Boecker is supported by the Wellcome Trust.

	FOOTNOTES

  Address for reprint requests: H. Boecker, Neurologische Klinik, Technische Universität München, Klinikum Rechts der Isar, Möhlstr. 28, D-81675 Munich, Germany.

  Received 4 June 1997; accepted in final form 8 October 1997.

	REFERENCES

Abstract Introduction Methods Results Discussion References

• ALBIN, R. L., YOUNG, A. B., PENNEY, J. B. The functional anatomy of basal ganglia disorders (with Comments). Trends Neurosci 12: 366-375, 1989.[Medline]
• ANDERSON, T. J., JENKINS, I. H., BROOKS, D. J., HAWKEN, M. B., FRACKOWIAK, R. S., KENNARD, C. Cortical control of saccades and fixation in man. A PET study. Brain 117: 1073-1084, 1994.[Abstract/Free Full Text]
• BAILEY, D. L., JONES, T., SPINKS, T. J., GILARDI, M.-C., TOWNSEND, D. W. Noise equivalent count measurements in a neuro-PET scanner with retractable septa. IEEE Trans. Med. Imaging 10: 256-260, 1991. [Medline]
• BENECKE, R., DICK, J. P., ROTHWELL, J. C., DAY, B. L., MARSDEN, C. D. Increase of the bereitschaftspotential in simultaneous and sequential movements. Neurosci. Lett. 62: 347-352, 1985.[Medline]
• BLINKENBERG, M., BONDE, C., HOLM, S., SVARER, C., ANDERSEN, J., PAULSON, O. B. Rate dependence of regional cerebral activation during performance of a repetitive motor task: a PET study. J. Cereb. Blood Flow Metab. 16: 794-803, 1996.[Medline]
• BOECKER, H., KLEINSCHMIDT, A., REQUARDT, M., HANICKE, W., MERBOLDT, K. D., FRAHM, J. Functional cooperativity of human cortical motor areas during self-paced simple finger movements. A high-resolution MRI study. Brain 117: 1231-1239, 1994.[Abstract/Free Full Text]
• BROOKS, D. J. The role of the basal ganglia in motor control: contributions from PET. J. Neurol. Sci. 128: 1-13, 1995.[Medline]
• BROTCHIE, P., IANSEK, R., HORNE, M. K. Motor function of the monkey globus pallidus. 2. Cognitive aspects of movement and phasic neuronal activity. Brain 114: 1685-1702, 1991.[Abstract/Free Full Text]
• CHEN, R., GERLOFF, C., HALLETT, M., COHEN, L. G. Involvement of the ipsilateral motor cortex in finger movements of different complexities. Ann. Neurol. 41: 247-254, 1997.[Medline]
• COLEBATCH, J. G., DEIBER, M. P., PASSINGHAM, R. E., FRISTON, K. J., FRACKOWIAK, R. S. Regional cerebral blood flow during voluntary arm and hand movements in human subjects. J. Neurophysiol. 65: 1392-1401, 1991.[Abstract/Free Full Text]
• DEECKE, L., SCHEID, P., KORNHUBER, H. H. Distribution of readiness potential, pre-motion positivity, and motor potential of the human cerebral cortex preceeding voluntary finger movements. Exp. Brain Res. 7: 158-168, 1969.[Medline]
• DEIBER, M.-P., IBAÑEZ, V., SADATO, N., HALLETT, M. Cerebral structures participating in motor preparation in humans: a positron emission tomography study. J. Neurophysiol. 75: 233-247, 1996.[Abstract/Free Full Text]
• DEIBER, M. P., PASSINGHAM, R. E., COLEBATCH, J. G., FRISTON, K. J., NIXON, P. D., FRACKOWIAK, R. S. Cortical areas and the selection of movement: a study with positron emission tomography. Exp. Brain Res. 84: 393-402, 1991.[Medline]
• DETTMERS, C., FINK, G. R., LEMON, R. N., STEPHAN, K. M., PASSINGHAM, R. E., SILBERSWEIG, D., HOLMES, A., RIDDING, M. C., BROOKS, D. J., FRACKOWIAK, R. S. Relation between cerebral activity and force in the motor areas of the human brain. J Neurophysiol. 74: 802-815, 1995.[Abstract/Free Full Text]
• FLAHERTY, A. W., GRAYBIEL, A. M. Input-output organization of the sensorimotor striatum in the squirrel monkey. J. Neurosci. 14: 599-610, 1994.[Abstract]
• FOX, P. T., FOX, J. M., RAICHLE, M. E., BURDE, R. M. The role of cerebral cortex in the generation of voluntary saccades: a positron emission tomographic study. J. Neurophysiol. 54: 348-369, 1985.[Abstract/Free Full Text]
• FRACKOWIAK, R. S., FRISTON, K. J. Functional neuroanatomy of the human brain: positron emission tomography-a new neuroanatomical technique. J. Anat. 184: 211-225, 1994.
• FREUND, H. J. Differential effects of cortical lesions in humans. Ciba Found. Symp. 132: 269-281, 1987.[Medline]
• FRISTON, K. J., ASHBURNER, J., POLINE, J.-B., FRITH, C. D., HEATHER, J. D., FRACKOWIAK, R.S.J. Spatial registration and normalization of images. Hum. Brain Mapp. 2: 1-25, 1995a.
• FRISTON, K. J., FRITH, C. D., LIDDLE, P. F., FRACKOWIAK, R. S. Comparing functional (PET) images: the assessment of significant change. J. Cereb. Blood Flow Metab. 11: 690-699, 1991.[Medline]
• FRISTON, K. J., FRITH, C. D., LIDDLE, P. F., FRACKOWIAK, R. S. Functional connectivity: the principal-component analysis of large (PET) data sets. J. Cereb. Blood Flow Metab. 13: 5-14, 1993.[Medline]
• FRISTON, K. J., HOLMES, A. P., WORSLEY, K. J., POLINE, J.-B., FRITH, C. D., FRACKOWIAK, R.S.J. Statistical parametric maps in functional imaging: a general linear approach. Hum. Brain Mapp. 2: 189-210, 1995b.
• FRISTON, K. J., WORSLEY, K. J., FRACKOWIAK, R.S.J., MAZZIOTTA, J. C., EVANS, A. C. Assessing the significance of focal activations using their spatial extent. Hum. Brain Mapp. 1: 214-220, 1994.
• GOLDBERG, G. Supplementary motor area structure and function: review and hypotheses. Behav. Brain Sci. 8: 567-616, 1985.
• GRAFTON, S. T., MAZZIOTTA, J. C., PRESTY, S., FRISTON, K. J., FRACKOWIAK, R. S., PHELPS, M. E. Functional anatomy of human procedural learning determined with regional cerebral blood flow and PET. J. Neurosci. 12: 2542-2548, 1992a.[Abstract]
• GRAFTON, S. T., MAZZIOTTA, J. C., WOODS, R. P., PHELPS, M. E. Human functional anatomy of visually guided finger movements. Brain 115: 565-587, 1992b.[Abstract/Free Full Text]
• HALSBAND, U., ITO, N., TANJI, J., FREUND, H. J. The role of premotor cortex and the supplementary motor area in the temporal control of movement in man. Brain 116: 243-266, 1993.[Abstract/Free Full Text]
• HALSBAND, U., MATSUZAKA, Y., TANJI, J. Neuronal activity in the primate supplementary, pre-supplementary and premotor cortex during externally and internally instructed sequential movements. Neurosci Res 20: 149-155, 1994.[Medline]
• IKEDA, A., LUDERS, H. O., BURGESS, R. C., SHIBASAKI, H. Movement-related potentials recorded from supplementary motor area and primary motor area. Role of supplementary motor area in voluntary movements. Brain 115: 1017-1043, 1992.[Abstract/Free Full Text]
• JAHANSHAHI, M., JENKINS, I. H., BROWN, R. G., MARSDEN, C. D., PASSINGHAM, R. E., BROOKS, D. J. Self-initiated versus externally triggered movements. I. An investigation using measurement of regional cerebral blood flow with PET and movement-related potentials in normal and Parkinson's disease subjects. Brain 118: 913-933, 1995.[Abstract/Free Full Text]
• JENKINS, I. H., BROOKS, D. J., NIXON, P. D., FRACKOWIAK, R. S., PASSINGHAM, R. E. Motor sequence learning: a study with positron emission tomography. J. Neurosci. 14: 3775-3790, 1994.[Abstract]
• JENKINS, I. H., PASSINGHAM, R. E., BROOKS, D. J. The effect of movement frequency on cerebral activation: a positron emission tomography study. J. Neurol. Sci. 151: 195-205, 1997.[Medline]
• JUEPTNER, M., STEPHAN, K.-M., FRITH, C. D., BROOKS, D. J., FRACKOWIAK, R.S.J., PASSINGHAM, R. E. Anatomy of motor learning: frontal cortex and attention to action. J. Neurophysiol. 77: 1313-1324, 1997.[Abstract/Free Full Text]
• KIM, S. G., ASHE, J., GEORGOPOULOS, A. P., MERKLE, H., ELLERMANN, J. M., MENON, R. S., OGAWA, S., UGURBIL, K. Functional imaging of human motor cortex at high magnetic field. J. Neurophysiol. 69: 297-302, 1993.[Abstract/Free Full Text]
• KITAMURA, J., SHIBASAKI, H., TAKAGI, A., NABESHIMA, H., YAMAGUCHI, A. Enhanced negative slope of cortical potentials before sequential as compared with simultaneous extensions of two fingers. Electroencephalogr. Clin. Neurophysiol. 86: 176-182, 1993.[Medline]
• LANG, W., LANG, M., UHL, F., KOSKA, C., KORNHUBER, A., DEECKE, L. Negative cortical DC shifts preceeding and accompanying simultaneous and sequential finger movements. Exp. Brain Res. 71: 579-587, 1988.[Medline]
• LUPPINO, G., MATELLI, M., CAMARDA, R., RIZZOLATTI, G. Corticocortical connections of area F3 (SMA-proper) and area F6 (pre-SMA) in the macaque monkey. J. Comp. Neurol. 338: 114-140, 1993.[Medline]
• LUPPINO, G., MATELLI, M., CAMARDA, R. M., GALLESE, V., RIZOLATTI, G. Multiple representations of bdy movements in mesial area 6 and the adjacent cingulate cortex: an intracortical microstimulation study in the macaque monkey. J. Comp. Neurol. 311: 463-482, 1991.[Medline]
• MARSDEN, C. D. What do the basal ganglia tell premotor cortical areas? In: Motor Areas of the Cerebral Cortex. Chichester, UK: Wiley, Ciba Foundation Symposium, 1987. p. 282-300.
• MARSDEN, C. D., DEECKE, L., FREUND, H. J., HALLETT, M., PASSINGHAM, R. E., SHIBASAKI, H., TANJI, J., WIESENDANGER, M. The functions of the supplementary motor area. Summary of a workshop. In: Advances in Neurology. Supplementary Sensorimotor area., edited by and H. O. Lüders . Philadelphia, PA: Lippincott-Raven, 1996, p. 477-487
• MARSDEN, C. D., OBESO, J. A. The functions of the basal ganglia and the paradox of stereotaxic surgery in Parkinson's disease (with Comments). Brain 117: 877-897, 1994.[Abstract/Free Full Text]
• MATELLI, M., LUPPINO, G., RIZZOLATTI, G. Architecture of superior and mesial area 6 and the adjacent cingulate cortex in the macaque monkey. J. Comp. Neurol. 311: 445-462, 1991.[Medline]
• MATELLI, M., RIZZOLATTI, G., BETTINARDI, V., GILARDI, M. C., PERANI, D., RIZZO, G., FAZIO, F. Activation of precentral and mesial motor areas during the execution of elementary proximal and distal arm movements: a PET study. Neuroreport 4: 1295-1298, 1993.[Medline]
• MATSUZAKA, Y., AIZAWA, H., 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.[Abstract/Free Full Text]
• MINK, J. W., THACH, W. T. Basal ganglia motor control. III. Pallidal ablation: normal reaction time, muscle cocontraction, and slow movement. J. Neurophysiol. 65: 330-351, 1991.[Abstract/Free Full Text]
• MUSHIAKE, H., INASE, M., TANJI, J. Selective coding of motor sequence in the supplementary motor area of the monkey cerebral cortex. Exp. Brain Res. 82: 208-210, 1990.[Medline]
• MUSHIAKE, H., INASE, M., TANJI, J. Neuronal activity in the primate premotor, supplementary, and precentral motor cortex during visually guided and internally determined sequential movements. J. Neurophysiol. 66: 705-718, 1991.[Abstract/Free Full Text]
• OLDFIELD, R. C. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 9: 97-113, 1971.[Medline]
• O'SULLIVAN, E. P., JENKINS, I. H., HENDERSON, L., KENNARD, C., BROOKS, D. J. The functional anatomy of remembered saccades: a PET study. Neuroreport 6: 2141-2144, 1995.[Medline]
• PASSINGHAM, R. E. Two cortical systems for directing movement. Ciba. Found. Symp. 132: 151-164, 1987.[Medline]
• PASSINGHAM, R. E. Functional specialization of the supplementary motor area in monkeys and humans. Adv. Neurol. 70: 105-116, 1996.[Medline]
• PETIT, L., ORSSAUD, C., TZOURIO, N., CRIVELLO, F., BERTHOZ, A., MAZOYER, B. Functional anatomy of a prelearned sequence of horizontal saccades in humans. J. Neurosci. 16: 3714-3726, 1996.[Abstract/Free Full Text]
• PETIT, L., ORSSAUD, C., TZOURIO, N., SALAMON, G., MAZOYER, B., BERTHOZ, A. PET study of voluntary saccadic eye movements in humans: basal ganglia-thalamocortical system and cingulate cortex involvement. J. Neurophysiol. 69: 1009-1017, 1993.[Abstract/Free Full Text]
• PICARD, N., STRICK, P. L. Motor areas of the medial wall: a review of their location and functional activation. Cereb. Cortex 6: 342-353, 1996.[Abstract/Free Full Text]
• PLAYFORD, E. D., JENKINS, I. H., PASSINGHAM, R. E., NUTT, J., FRACKOWIAK, R. S., BROOKS, D. J. Impaired mesial frontal and putamen activation in Parkinson's disease: a positron emission tomography study. Ann. Neurol. 32: 151-161, 1992.[Medline]
• RAO, S. M., BANDETTINI, P. A., BINDER, J. R., BOBHOLZ, J. A., HAMMEKE, T. A., STEIN, E. A., HYDE, J. S. Relationship between finger movement rate and functional magnetic resonance signal change in human primary motor cortex. J. Cereb. Blood Flow Metab. 16: 1250-1254, 1996.[Medline]
• RAO, S. M., BINDER, J. R., BANDETTINI, P. A., HAMMEKE, T. A., YETKIN, F. Z., JESMANOWICZ, A., LISK, L. M., MORRIS, G. L., MUELLER, W. M., ESTKOWSKI, L. D., WONG, E. C., HAUGHTON, V. M., HYDE, J. S. Functional magnetic resonance imaging of complex human movements. Neurology 43: 2311-2318, 1993.[Abstract/Free Full Text]
• REMY, P., ZILBOVICIUS, M., LEROY WILLIG, A., SYROTA, A., SAMSON, Y. Movement- and task-related activations of motor cortical areas: a positron emission tomographic study. Ann. Neurol. 36: 19-26, 1994.[Medline]
• RIZZOLATTI, G., LUPPINO, G., MATELLI, M. The classic supplementary motor area is formed by two independent areas. Adv. Neurol. 70: 45-56, 1996.[Medline]
• ROLAND, P. E. Organization of motor control by the normal human brain. Hum. Neurobiol. 2: 205-216, 1984.[Medline]
• ROLAND, P. E. Cortical organization of voluntary behavior in man. Hum. Neurobiol. 4: 155-167, 1985.[Medline]
• ROLAND, P. E., LARSEN, B., LASSEN, N. A., SKINHOJ, E. Supplementary motor area and other cortical areas in organization of voluntary movements in man. J. Neurophysiol. 43: 118-136, 1980a.[Abstract/Free Full Text]
• ROLAND, P. E., MEYER, E., SHIBASAKI, T., YAMAMOTO, Y. L., THOMPSON, C. J. Regional cerebral blood flow changes in cortex and basal ganglia during voluntary movements in normal human volunteers. J. Neurophysiol. 48: 467-480, 1982.[Free Full Text]
• ROLAND, P. E., SKINHOJ, E., LASSEN, N. A., LARSEN, B. Different cortical areas in man in organization of voluntary movements in extrapersonal space. J. Neurophysiol. 43: 137-150, 1980b.[Abstract/Free Full Text]
• SADATO, N., CAMPBELL, G., IBANEZ, V., DEIBER, M., HALLETT, M. Complexity affects regional cerebral blood flow change during sequential finger movements. J. Neurosci. 16: 2691-2700, 1996a.[Abstract]
• SADATO, N., IBANEZ, V., DEIBER, M. P., CAMPBELL, G., LEONARDO, M., HALLETT, M. Frequency-dependent changes of regional cerebral blood flow during finger movements. J. Cereb. Blood Flow Metab. 16: 23-33, 1996b.[Medline]
• SERGENT, J., ZUCK, E., TERRIAH, S., MACDONALD, B. Distributed neural network underlying musical sight-reading and keyboard performance. Science 257: 106-109, 1992.[Abstract/Free Full Text]
• SHIBASAKI, H., SADATO, N., LYSHKOW, H., YONEKURA, Y., HONDA, M., NAGAMINE, T., SUWAZONO, S., MAGATA, Y., IKEDA, A., MIYAZAKI, M., FUKUYAMA, H., ASATO, R., KONISHI, J. Both primary motor cortex and supplementary motor area play an important role in complex finger movement. Brain 116: 1387-1398, 1993.[Abstract/Free Full Text]
• SILBERSWEIG, D. A., STERN, E., FRITH, C. D., CAHILL, C., SCHNORR, L., GROOTOONK, S., SPINKS, T., CLARK, J., FRACKOWIAK, R., JONES, T. Detection of thirty-second cognitive activations in single subjects with positron emission tomography: a new low-dose H₂(¹⁵)O regional cerebral blood flow three-dimensional imaging technique. J Cereb. Blood Flow Metab. 13: 617-629, 1993.[Medline]
• SPINKS, T. J., JONES, T., BAILEY, D. L., TOWNSEND, D. W., GROOTOONK, S., BLOOMFIELD, P. M., GILARDI, M.-C., CASEY, M. E., SIPE, B., REED, J. Physical performance of a positron tomograph for brain imaging with retractable septa. Phys. Med. Biol. 37: 1637-1655, 1992.[Medline]
• STEPHAN, K. M., FINK, G. R., PASSINGHAM, R. E., SILBERSWEIG, D., CEBALLOS BAUMANN, A. O., FRITH, C. D., FRACKOWIAK, R. S. Functional anatomy of the mental representation of upper extremity movements in healthy subjects. J. Neurophysiol. 73: 373-386, 1995.[Abstract/Free Full Text]
• TALAIRACH, J., TOURNOUX, P. In: A Co-Planar Stereotaxic Atlas of a Human Brain., . Stuttgart, Germany: Thieme, 1988
• TANJI, J. The supplementary motor area in the cerebral cortex. Neurosci. Res. 19: 251-268, 1994.[Medline]
• TANJI, J., KURATA, K. Comparison of movement-related activity in two cortical motor areas of primates. J. Neurophysiol. 48: 633-653, 1982.[Free Full Text]
• TANJI, J., SHIMA, K. Role for supplementary motor area cells in planning several movements ahead. Nature 371: 413-416, 1994.[Medline]
• TANJI, J., SHIMA, K. Contrast of neuronal activity between the supplementary motor area and other cortical motor areas. In: Advances in Neurology. Supplementary Sensorimotor Area, edited by and H. O. Lüders . Philadelphia, PA: Lippincott-Raven, 1996, p. 95-103
• TYSZKA, J. M., GRAFTON, S. T., CHEW, W., WOODS, R. P., COLLETTI, P. M. Parceling of mesial frontal motor areas during ideation and movement using functional magnetic resonance imaging at 1.5 tesla. Ann. Neurol. 35: 746-749, 1994.[Medline]
• WESSEL, K., ZEFFIRO, T., LOU, J. S., TORO, C., HALLETT, M. Regional cerebral blood flow during a self-paced sequential finger opposition task in patients with cerebellar degeneration. Brain 118: 379-393, 1995.[Abstract/Free Full Text]
• WORSLEY, K. J., EVANS, A. C., MARRETT, S., NEELIN, P. J. A three-dimensional statistical analysis for rCBF activation studies in human brain. J. Cereb. Blood Flow Metab. 12: 900-918, 1992.[Medline]
• ZILLES, K., SCHLAUG, G., GEYER, S., LUPPINO, G., MATELLI, M., QU, M., SCHLEICHER, A., SCHORMANN, T. Anatomy and transmitter receptors of the supplementary motor areas in the human and nonhuman primate brain. Adv. Neurol. 70: 29-43, 1996.[Medline]
• ZILLES, K., SCHLAUG, G., MATELLI, M., LUPPINO, G., SCHLEICHER, A., QU, M., DABRINGHAUS, A., SEITZ, R., ROLAND, P. E. Mapping of human and macaque sensorimotor areas by integrating architectonic, transmitter receptor, MRI and PET data. J. Anat. 187: 515-537, 1995.

0022-3077/98 $5.00 Copyright ©1998 The American Physiological Society