GENERATION AND CHARACTERIZATION OF CORRELATED SPIKE TRAINS

A framework within which to generate and characterize populations of c orrelated spike trains is presented. Spike trains are generated using integrate to threshold and fire type encoders. Spectral analysis techn iques form the basis for specifying the strength of correlation within the population. To accurately specify weak correlation, a combined co herence estimate is formed between several independent pairs of spike trains from a population. The aim is to produce populations of spike t rains which have a realistic stochastic correlation structure and whic h can be used to explore the role of correlated neuronal discharge in information processing in neuronal systems. (C) 1998 Elsevier Science Ltd. All rights reserved.