NEURONAL MODELING IN THE PRESENCE OF RANDOM REFRACTORINESS

Abstract

Inclusion of refractoriness in a diffusion model for single neuron activity is the object of the present paper. An asymptotic analysis of the random process modeling the number of firings and the distribution of interspike intervals is performed under the approximation of exponentially distributed firings. In the cases of constant, uniform, exponential, Erlang, truncated normal and hyperexponential durations of refractory periods, asymptotic simple expressions are obtained and some numerical evaluations are performed