Motion Selectivity of the Local Filed Potentials in the Primary Visual Cortex of Rats: A Machine Learning Approach

Abstract

Using rodents as a model of physiological vision studies requires adequate information about their visual cortex. Although the primary visual cortex of rats has different sub-regions, there are few studies on the different response patterns of these sub-regions. In this study, we recorded the local field potentials (LFPs) from sub-regions of the primary visual cortex (V1) of anesthetized rats. We used random dots patterns as moving stimuli presented in random sequences. Then we used machine learning methods to decode the direction and speed of the stimuli from the recorded signals. Our results revealed that there are different patterns of responses to motion stimuli across sub-regions. Although the decoding results using LFPs were not high, they were enhanced by moving to the lateral sub-regions of the V1. Our results suggested that the location of the recording areas impact reaction time, the pattern of the responses in time- and frequency- domains, and encoding the motion stimuli.