Poster Abstract

May 14, 2010

A Computational Study on the Effect of Firing Rate Properties on Multi-Unit Information

Corey Baker
LSAMP VI Bridges to the Doctorate Scholar

Abstract: I worked with Professor Deborah Won on a project at CSULA in the area of spike sorting in brain machine interfaces (BMI). Spike sorting is often used in BMI, but the ability to discriminate neural activity is extremely complicated and imposes a significant computational burden. The research compared information in pooled (non-sorted) and label-lined (sorted) populations of firing rates of neurons. I examined the effects of tuning width on the differences in information. Neurons were tuned to fire at different arm movement directions. The tuning widths of the neural activity were varied by using Von-Mises tuning curves. The research culminated in a paper that discussed the conditions under which spike sorting may or may not be needed