Finding Reliable Activation of Neural Sequences in the Visual Cortex of Macaque Monkeys

BIOE 313: Neuromorphics: Brains in Silicon

Stanford Bioengineering, March - June 2021

Abstract

Based on the paper I presented in class, Reliable sequential activation of neural assemblies by single pyramidal cells in a three-layered cortex by Gilles Laurent et al., I use a publicly available dataset to find evidence for neurons firing in specific sequences dependent on the stimuli presented. There are many publicly available datasets that contain neural signals, including Neurodata Without Borders (NWB) and Collaborative Research in Computational Neuroscience (CRCNS). I use recordings of single neurons taken using a Utah Multi-Electrode Array placed in the visual cortex of anesthetized macaque monkeys from CRCNS [2]. In the paper presented by Laurent et al., they stimulated single pyramidal neurons in layer 3 of turtle cortices and recorded the response of the surrounding neurons. The data I use involves a visual stimulus being presented to the monkey in the form of a flash on a screen. Rather than a single pyramidal neuron being stimulated, this experiment involves the activity of many neurons in the primary visual cortex due to the visual stimuli presented to the monkeys. The data includes ten recording sessions from three animals with arrays of spike trains from 207 neurons in the format [#neurons #image #trials #milliseconds]. Example images can be seen in figure 1. Using this data, I identify neurons who increase firing rate by 1 standard deviation above baseline firing rate after stimulus is presented (followers). Based on the firing-rate onset time, I rank order each follower and identify reliable sequences that occur when visual stimuli are presented.