Universal Transient Dynamics in Oscillatory Network Models of Epileptic Seizures

Abstract

Discharges of different epilepsies are characterized by different signal shape and duration. The authors adhere to the hypothesis that spike-wave discharges are long transient processes rather than attractors. This helps to explain some experimentally observed properties of discharges, including the absence of a special termination mechanism and quasi-regularity. Analytical approaches mostly cannot be applied to studying transient dynamics in large networks. Therefore, to test the observed phenomena for universality one has to show that the same results can be achieved using different model types for nodes and different connectivity terms. Here, we study a class of simple network models of a thalamocortical system and show that for the same connectivity matrices long, but finite in time quasi-regular processes mimicking epileptic spike-wave discharges can be found using nodes described by three neuron models: FitzHugh – Nagumo, Morris – Lecar and Hodgkin – Huxley. This result takes place both for linear and nonlinear sigmoid coupling.