The activity of interneurons in the spinal cord is dominated by distributed dynamic patterns that are tightly coordinated across the population. These distributed patterns are consistent with an increasingly popular state space perspective on neural population activity which posits that a coordinated dynamic system underlies the activity of individual neurons and the state of the network and its temporal evolution are the key variables that determine the circuit’s function. However neural population states and dynamics are often interrogated at course timescales that do not match the temporal precision at which neural circuits operate. Here we used artificial intelligence methods to uncover spinal interneuron population dynamics with high precision on the time scale of milliseconds. Analysis of the uncovered representations revealed precise links between the dominant patterns of spinal interneuron population activity and the muscle activity output from the spinal cord. In particular we observed a regional organization in interneuron state space i.e. transitions in and out of specific state space regions were linked to transitions between flexor and extensor activation precise on the order of milliseconds. Additionally we found a precise correspondence between interneuron state space trajectories and single-cycle variations in the magnitude of muscle activations. Our results demonstrate the power of AI methods for precisely revealing the link between neural population dynamics and circuit function.
Biography: Chethan Pandarinath is a tenured Associate Professor in the Coulter Department of Biomedical Engineering and Neurosurgery at Emory University and Georgia Tech. He directs the Systems Neural Engineering Lab whose research sits at the intersection of neural engineering systems neuroscience and artificial intelligence with dual goals of better understanding the nervous system and designing assistive devices for people with paralysis. Dr. Pandarinath is a 2019 Sloan Fellow a recipient of the 2021 NIH Director’s New Innovator Award and a 2024 Gilbreth Lecture at the National Academy of Engineering. His work has been funded by the Neilsen Foundation NSF DARPA Air Force Burroughs Wellcome Fund Simons Foundation and NIH.