Sparsity and Complexity in V1 Neuronal Codes Revealed by Large-scale Long-term Two-photon Calcium Imaging
Abstract:
Neurophysiological investigations of how the primate cortex represents the visual world are limited by the number of neurons that can be monitored simultaneously and the number of stimuli that can be tested. Here, we establish the feasibility of large-scale long-term imaging of neurons in the primary visual cortex (V1) of awake macaques. By using two-photon microscopy with genetically-encoded calcium-indicator, we could simultaneously image thousands of neurons and systematically characterize their responses to over 10,000 visual stimuli, ranging from standard oriented bars, edges, and gratings to more complex patterns and natural images. We found that only 0.5% of the V1 neurons in the superficial layer responded to a given stimulus, with individual neurons characteristically responding only to 0.5% of the presented natural images. Such sparsity in V1 neuronal responses is an order of magnitude higher than the earlier estimates. Most neurons responded to complex patterns as a whole, rather than simple patterns. The observed high degree of response sparsity in V1 superficial layer neurons, with their stimulus preferences and specificities to complex patterns, provides a new perspective on V1 neuronal codes.
Host: Tai Sing Lee.
P.S. Dr. Tang is on a tight schedule and will be available for meeting only in the morning 9:00-11:00 a.m. and early afternoon (1:30-3:30) of October 22. If you wish to meet with him, please contact Maureen Kelly ( makelly@andrew.cmu.edu) and cc me at tai@cnb.cmu.edu .