Interspike interval statistics obtained from non-homogeneous gamma spike generator

Abstract

The lengths of interspike intervals between two successive spikes in a neural spike train often vary both within and across trials. In order to describe and analyze neuronal firing, statistical models and methods of probability theory and stochastic point process have been widely applied. In this study, we compare several non-homogeneous gamma processes on the basis of reproducing the wide distribution of the irregularity statistics in vivo and show some conditions for reproducibility. We conclude that the changes of firing rates are not sufficient for describing the fluctuations of the statistics. © 2009 Springer-Verlag Berlin Heidelberg.