Population coding of movement dynamics by cerebellar Purkinje cells

In order to accomplish desired movements, the nervous system must specify the movement dynamics: it must provide a signal that compensates for the mechanical constraints encountered during movement. Here we tested whether the population activity of Purkinje cells in the cerebellum specifies the dynamics for pursuit eye movements. We first estimated the population activity by computing weighted averages of Purkinje cell firing on a millisecond time scale. We then generated predicted eye movements by transforming this pooled neural activity with a description of eye mechanics. We found that the equally weighted average of Purkinje cell outputs produced a close match between the predicted and actual eye movements. These findings demonstrate that neural circuits through the cerebellum are capable of providing the dynamic compensation necessary to achieve desired movements.