Function of Cerebellar Microcircuitry within a Closed-loop System during Control and Adaptation

Abstract: The human motor control is both robust and stable, despite large delays and highly complex motor systems with an abundance of actuators, sensors and degrees of freedom. The cerebellum is thought help accomplish this by compensating for external loads and internal limitations and disturbances through adaptation, creating inverse models of the motor system dynamics. The cerebellum does also exhibit a generic relatively well described modular microcircuitry, making it a suitable neural circuitry to study. This thesis models a small part of the cerebellum, using detailed bio-physical models in combination with rate-based models, and uses the constructed network model to improve control of a planar double joint arm.The individual neuron models were calibrated using data from in vivo experiments. The response from the models when they were introduced to recorded primary afferent spike trains, originating from tactile stimulation, was used to validate their behaviour. Subsets of the complete network was also constructed to investigate possible functions of the granule cells and inhibitory connection patterns between interneurons within the molecular layer.