Source location as a feature for the classification of multi-sensor extracellular action potentials

Extracellular action potentials (EAPs) must be classified before they can yield any useful information on neuronal function and organization. Neuronal source classification therefore represents a critical step in the analysis of electrophysiological data. This study demonstrates the efficacy of a multi-sensor EAP classification scheme using source location as a classification feature. Localization was performed using the multiple signal classification (MUSIC) algorithm. Six distinct source neurons were classified from 20 seconds of extracellular, four-sensor (tetrode) recordings. On average, 89.5% of the waveforms making up each class matched the shape of the average class waveform. These results indicate that this classification scheme can successfully identify individual neurons from multi-sensor EAP recordings.