Analysis of spike propagation in feed-forward neural networks

Abstract

Propagation of spiking activity in networks of neuron is the key to how information is presented in the brain. The precise arrival times of spikes encode different kinds of incoming information. It has been shown that parameters like the number of spikes and the variability in timing of spikes determine the propagation of input across the network (Diesmann, Gewaltig, & Aertsen,1999b). We want to determine the conditions under which the activity propagates across layers of neurons in a feed-forward network (Diesmann, Gewaltig, & Aertsen, 1999a; Kumar, Rotter, & Aertsen, 2010a; Rossum,Turrigiano, & Nelson, 2002) Also we look for the conditions under which spike volleys with spread in spike times synchronise and vice-versa. Feed-forward networks are embedded in many brain areas (Doupe, Solis, Kimpo, & Boettiger, 2004; Fee & Scharff, 1969; Gewaltig, Diesmann, & Aertsen, 2001a; Hanuschkin, Diesmann, & Morrison, 2011; Mittmann, Koch, & Häusser, 2005; Teramae & Fukai, 2008) and an understanding of activity propagation in canonical feed-forward network will help us understand dynamics of neural networks in general and also form a good framework for understanding sensory computation.