Basal ganglia (BG) constitute a network of 7 deep brain nuclei involved in a variety of crucial brain functions including: action selection, action gating, reward based learning, motor preparation, timing etc. In spite of the immense amount of data available today, researchers continue to wonder how such a small deep brain circuit performs such a bewildering range of functions. Computational models of BG have focused on individual functions and fail to give an integrative picture of BG function. A major breakthrough in our understanding of BG function is perhaps the insight that activities of mesencephalic dopaminergic cells represent some form of ‘reward’ to the organism. This insight enabled application of tools from ‘reinforcement learning’ (RL), a branch of machine learning, in the study of BG function. Of the key pathways that constitute the BG circuit - the Direct and the Indirect Pathways - the function of the latter pathway has been given a variable and tentative interpretation. Exploration, a key RL mechanism, is shown to have a cortical substrate but no counterpart in BG circuit has been found. We hypothesize that the STN-GPe part of BG is the Explorer and show using simulations, how such an assumption completes the picture. Presenting a series of models describing how BG is involved in control of spatial navigation, reaching, handwriting, saccade generation etc., we outline an approach towards development of an integrated computational model of Basal Ganglia.

　大脳基底核は、神経生理学、神経解剖学、臨床神経学ばかりでなく、数理工学からも注目を集めており、ホットな分野の一つです。様々な数理モデルが 提唱され、それによって大脳基底核の機能の一端が明らかになりつつあります。Chakravarthy博士は、現在、理研BSIの甘利研に滞在しており、 今回、生理研を訪問されるので、セミナーをお願いしました。是非、御来聴下さい。