Home Research BCCN I Project Details A2 - A theory for the origin of patterns of precisely timed spikes
Personal tools

A2 - A theory for the origin of patterns of precisely timed spikes

Marc Timme

Repeated, precisely timed spike patterns have been observed experimentally in different neuronal systems and are discussed to be an essential part of neural information processing (Abeles et al. 1993, Nadasdy et al. 1996, Diesmann et al. 1999). Their dynamical origin, however, remains unknown. Theoretical modeling studies have shown that periodic spike patterns can be attractors of the nonlinear dynamics of neural networks (Jin 2002). Using analytical methods recently developed to study the dynamics of networks with a complicated, biologically realistic connectivity (Timme et al. 2002b), we have now demonstrated how network connectivity and heterogeneity determine the precise timing of spikes (Denker et al. 2004). In contrast to a common hypothesis that precisely timed spike patterns are generated in excitatorily coupled feed-forward structures, our analytically tractable model demonstrates that recurrent inhibition provides another mechanism for creating spike patterns. Furthermore our analysis shows that the temporal spread of a pattern that emerges from synchrony due to heterogeneous couplings is bounded above by the delay time of the interactions. An open key question is how spike patterns of realistic temporal extent can emerge. Our previous results (Denker et al. 2004), based on static synapses, have shown that the details of the synaptic connectivity, or, more generally, the structure of the underlying network has a major effect on the spike pattern exhibited.

Belongs to Group(s):