Informational Estimates of Neuronal Responses

Abstract

We studied responses of visually driven neurons from the associative Clare - Bishop area of the cat cortex. Interspike intervals of neuronal responses were subdivided into four constituents: intervals between all spikes, intervals between spikes inside bursts, intervals between bursts, and intervals between events (in which case it has been assumed that a burst and an isolated spike are events of equal importance). Two informational parameters based on these distributions were calculated: the information capacity of the channel (F) and the amount of information (entropy H). We found that (1) values for F and H are significantly higher when they are based on intervals between all spikes and intervals between events than when they are based on intervals inside bursts and intervals between bursts; (2) values of F and H are significantly higher for all cases where a neuronal response is present than those for spontaneous activity.