An Implementation of a Spike-Response Model With Escape Noise Using an Avalanche Diode

Abstract

This paper introduces a novel probabilistic spike-response model through the combination of avalanche diode-generated Poisson distributed noise, and a standard exponential decay-based spike-response curve. The noise source, which is derived from a 0.35-μm single-photon avalanche diode (kept in the dark), was tested experimentally to verify its characteristics, before being combined with a field-programmable gate-array implementation of a spike-response model. This simple model was then analyzed, and shown to reproduce seven of eight behaviors recorded during an extensive study of the ventral medial hypothalamic (VMH) region of the brain. It is thought that many of the cell types found within the VMH are fed from a tonic noise synaptic input, where the patterns generated are a product of their spike response and not their interconnection. This paper shows how this tonic noise source can be modelled, and due to the independent nature of the noise sources, provides an avenue for the exploration of networks of noise-fueled neurons, which play a significant role in pattern generation within the brain.