Simulating Parkinson's disease patient deficits using a COVIS-based computational model

COVIS is a neurobiologically motivated model of perceptual category learning. It includes two competing systems: the hypothesis-testing system mediates learning and performance in tasks requiring explicit reasoning; the procedural system mediates learning and performance in tasks that are achieved procedurally through trial and error learning when no explicit rule/strategy exists. Here we describe a computational implementation of COVIS used to model the differential effects of dopamine depletion on performance in a perceptual category-learning task and the simplified Wisconsin Card Sorting Test (WCST).