Functional Role of the Basal Ganglia in the Planning and Execution of Actions Oury Monchi, PhD, 1,2 Michael Petrides, PhD, 3 Antonio P. Strafella, MD, PhD, 3 Keith J. Worsley, PhD, 3 and Julien Doyon, PhD 1,3,4 Objective: Recent studies of functional brain imaging have shown the involvement of the basal ganglia in executive processes such as planning and set-shifting. However, the specific contributions of the striatum in those processes remain unknown. This study aimed to test the hypothesis that the caudate nucleus is primarily involved in the preparation of a novel action and not in set-shifting per se. Methods: In the present event-related functional magnetic resonance imaging (fMRI) study, a new task was developed that permitted, for the first time, to distinguish between shifts in classification when the rule is implicitly given by the task from shifts that require cognitive comparison and planning. Results: Sig- nificantly increased activity in the caudate nucleus and the putamen was observed only in conditions in which cognitive planning was required to perform a set-shift, whereas significant activation was seen in the subthalamic nucleus (another region of the basal ganglia) in all shifting conditions whether or not planning was required. Interpretation: We suggest that the caudate nucleus and the putamen are particularly important, respectively, in the planning and the execution of a self-generated novel action, whereas the subthalamic nucleus may be required when a new motor program is solicited independently of the choice of strategy. Ann Neurol 2006;59:257–264 Abnormalities of the basal ganglia (BG) have been de- scribed in many neurological disorders such as Parkin- son’s disease, and thus it is important to gain a greater understanding of the functional role of the different nuclei of the BG in the healthy brain in order to reveal the neural origins of the cognitive and motor deficits observed in those disorders. Traditionally thought to be associated with the control of movement 1–3 and motor learning, 4,5 BG has been shown by neuropsy- chological studies in patients with Parkinson’s and Huntington’s disease to contribute to a variety of ex- ecutive functions, including planning and set-shifting (ie, the ability to alter our mode of response in the face of changing circumstances). 6–9 Among the different BG nuclei, the caudate nucleus has been shown to play a greater role in executive processes, 10 –14 whereas other structures such as the putamen and subthalamic nu- cleus have traditionally been associated with more motor-related activities. However, there is evidence that the role of the putamen may not be directly linked to the movement itself, but rather to the condition un- der which it is made, 15 whereas the subthalamic nu- cleus may exert a specific influence on the BG output related to the control of movement. 3,16,17 Hence, whereas the evidence accumulated to date suggests that the various nuclei of the BG may have different func- tional roles, their distinct contribution to the planning and execution of action remains unknown. Recently, functional magnetic resonance imaging (fMRI) has been used routinely to investigate the neu- ral substrates of various cognitive functions in normal subjects as well as patients with neurological or psychi- atric disorders. In the earlier fMRI studies, it was dif- ficult to make fine distinctions in cognitive processing and to distinguish between small neural structures, es- pecially subcortical ones. However, recent methodolog- ical advances both in the type of acquisition sequences and in data analysis has made it possible to study the involvement of subcortical structures, such as the dif- ferent parts of the BG, in specific cognitive conditions. In a previous event-related fMRI study of the Wiscon- sin Card-Sorting Task in young healthy adults, 12 we observed significant activation in the caudate nucleus specifically when subjects received negative feedback (ie, when a set-shift was required). While the involve- ment of the caudate nucleus in set-shifting has also been confirmed in other neuroimaging studies, 11,13 there is still no evidence up to now of whether the From the 1 Functional Neuroimaging Unit, Institut Universitaire de Ge ´riatrie de Montre ´al; 2 Department of Radiology, University of Montreal; 3 Montreal Neurological Institute, McGill University; and 4 Department of Psychology, University of Montreal, Montreal, Quebec, Canada. Received Jul 15, 2005, and in revised form Aug 30. Accepted for publication Oct 7, 2005. Published online Jan 23, 2006 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/ana.20742 Address correspondence to Dr Monchi, Centre de Recherhe, Insti- tut Universitaire de Ge ´riatrie de Montre ´al, 4565, Queen Mary Road, Montre ´al (Que ´bec) H3W 1W5, Canada. E-mail: oury.monchi@umontreal.ca or oury@bic.mni.mcgill.ca © 2006 American Neurological Association 257 Published by Wiley-Liss, Inc., through Wiley Subscription Services