Our work focuses on how the brain decides which action to perform at each moment in time – that is, action selection. We are interested in the cortical and subcortical circuits that mediate this process, and how they go awry in neurological disorders such as Parkinson’s disease. Specifically, we perform measurements of large-populations of neural activity in freely behaving mice using imaging and physiology, and distill their behavior in real-time using 3D cameras and probabilistic approaches to machine learning. Additionally, we are pursuing new methods to control activity in these circuits using precision closed-loop deep brain stimulation.