Neural Population Dynamics

Whether it's picking up a cup or deftly slotting a key into a lock, we move our arms and hands with ease. Yet behind that apparent ease lies some of the most complex computation the brain performs, and skilled control of the arm and hand is detrimentally affected by many diseases and injuries. We still lack a clear understanding of how neural networks in the brain control these skilled behaviours.

Understanding how these networks control reaching and grasping requires monitoring the activity of large neural populations during skilled behaviour. We use high-density probes to record simultaneously from multiple cortical and subcortical regions while non-human primates perform skilled reaching and grasping, and we record from populations of motor units in humans and non-human primates to investigate how muscles are coordinated during movement. This approach has shown, for example, that what we expect to sense reconfigures neural population dynamics in motor circuits, and that reward gates how subcortical structures encode the patterns we hear. We are now extending these techniques to record from the same populations over many weeks — a crucial step toward tracking how populations of neurons learn and flexibly adapt to new complex motor behaviours.