Detection of extracellular action potentials in noise for the control of microelectrode advancement

Abstract

A digital computer was programmed to detect impulses in the presence of noise, rather than identify or classify impulse activity from microelectrodes. The analog signal was abstracted into a sequential series of voltage time vectors that measured peak-to-peak activity. The amplitude and time difference between a peak-positive potential and the next peak-negative potential defined one vector. The amplitude and time difference between that negative peak and the next positive peak defined the next vector, and so on. An algorithm determined if each successive vector was part of a signal pattern by comparing the properties of the vector to those in a stored list. The algorithm was designed for future application with minimum computer systems and multiple-tip microelectrodes.