Neurocomputational modeling of the basal ganglia in motor learning at mesoscopic scale: an overview

Abstract

The first model of the basal ganglia (BG) was conceived almost half a century ago. Since then, extensive research efforts have been carried out to further refine and understand the physiological and pathological BG behaviour and role. Currently, it is well-known that the BG are crucial in motor learning and motor diseases are associated to dysfunction of the nuclei, such as the parkinsonian syndrome, dystonia, chorea, etc. We are still a long way from giving an answer to all the questions, but advances in technology are making research advance significantly in recent years. Computational modeling is one of these methodologies and allows to evaluate the interactions within and among multiple neural systems. The development and analysis of the behaviour of computational models, in concert with multimodal analysis and in vivo experiments, leads to new scientific results. This review provides a critical synopsis of the evolution of thought regarding the physiological model of the BG with respect to motor learning, revisiting past theories and summarizing the main recent findings to this field of research, to highlight their innovative contribution to knowledge of the functioning of the nuclei and formulate a state-of-art hypothesis of BG modeling.