Criteria for evaluating multiunit spike separation techniques

摘要

Multi-unit neural recordings often occur in neurophysiological studies, and a wide variety of techniques have been proposed to separate the multi-unit activity. In this paper we propose a method for evaluating the power of various multi-unit sorting techniques. The analysis can also be applied to a data set to determine whether a given multi-unit isolation technique can isolate a given set of wavelets to a specified statistical standard. The apparent separation matrix described in the paper is then used to evaluate the ability of four multi-unit isolation techniques to resolve a population of 8 units simultaneously recorded in the medial articular nerve of the cat.