Impaired representation of temporal statistics in cerebellar degeneration

Human cerebellar degeneration impairs Bayesian interval timing, showing greater timing variability, stronger regression toward the mean, and slower prior updating in duration reproduction tasks.

• Why it matters: The work addresses how sensory evidence is combined with prior knowledge in cerebellar circuits, a gap in understanding Bayesian inference for interval timing.
• What they did: Patients with cerebellar degeneration and matched controls performed duration reproduction tasks with varying prior distributions, and a cerebellar microcircuit model with Purkinje cell plasticity over a granule cell temporal basis set was fit to behavior.
• The result: Model simulations of cerebellar degeneration produced an output approximated a Bayesian observer with an intact prior but a noisier likelihood, matching patients’ behavior across a wide parameter space.

AI summary of the abstract · may contain errors · not medical advice · read the paper.