CRCNS: Computational Integration of Human Adaptation and Primate Neurophysiology

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): The neural control of movement encompasses several conundrums in computational neuroscience. The fundamental function of cerebellum and primary and premotor cortices remains elusive. We have recently pioneered behavioral experiments of haptic learning that identify the transformation from sensed movements to incremental adaptations. Simulations have predicted neuronal activities necessary to generate these transformations and have suggested surprising plasticity in the neural representation of movement. Permanent motor impairments due to degeneration, strokes, and traumas affect hundreds of thousands of individuals each year, rendering people unable to perform motor behaviors. A quantitative understanding of normal motor adaptation will enhance the ability of rehabilitation to help patients regain normal function. Here we propose serial human and primate experiments of learning novel visuomotor environments. Humans will experience virtual reality environments that alter the dependence of visual feedback on hand position; the transformation complexity will vary across training days. We will identify how people learn visuomotor transformations over sessions and trial-by-trial, enabling neuronal network simulations to predict neuronal activity needed to mimic human adaptation. We will then record in motor cortex, premotor cortex, and cerebellum as monkeys perform the same visuomotor adaptations, to determine how neuronal activities depend on the environment and on trial-by-trial adaptation. The within-session and across-session changes will test the network simulations and will elucidate the particular contribution of each cortical area as inputs of motor plans, adaptive transformations between vision and action, or outputs of generated movements. The proposed collaboration between Dr. Thoroughman, a computational neuroscientist and psychophysicist, and Dr. Moran, a primate neurophysiologist, will enable a rich connection between neural computation, adaptive behavior, and cortical activity. The new collaboration will directly impact the quality and specificity of leading theories of motor control and learning. We aim to formulate basic scientific foundations that will ultimately improve metrics of neurological diagnosis, inform the design of patientspecific therapies and neuro-rehabilitation protocols, and help patients generalize beyond clinical training to improve motor function in their daily lives.