Item

Abstract

Understanding curcuit computation in the nervous system requires sampling activity over large neural populations and maximizing the number of features that can be extracted . By combining planar arrays of extracullular electrodes with the three-layered cortex of turtles, we show that synaptic signals induced along individual axons as well as action poteintials can be easily captured. Two types of information can be extracted from these signals, the neuronal subtype (inhibitory or excitatory)-whose identification is more reliable than with traditional measures such as action potential width - and a (partial) spatial map of functional axonal projections from individual neurons. Because our approach is algorhytmic, it can be carried out in parellel on hundreds of simultaneoulsy recoreded neurons. Combining our approach with soma triangulation, we reveal an axonal projectionbias among a population of pyramidal neurons in turtle cortex and confirm this bias through anatomical reconstructions.