ABSTRACT
It is an unchallenged viewpoint that the collective behavior of a complex dynamical network like the brain emerges from both the local dynamics of the constituents and from the structure of the network that connects them (Strogatz 2001). Epileptic seizures are generally considered as a short episode of overly synchronized –ring of neurons. It could advance our understanding of ictogenesis to identify combinations of neuron and network properties that lead to synchronized seizurelike –ring, to asynchronous –ring, or to a seizure-prone con–guration with intermittent changes between these two dynamic states (Wendling 2005; Soltesz and Staley 2008; Lytton 2008). We aim at developing and studying such neuron network models that can spontaneously (i.e., without change of parameters) switch between asynchronous and synchronous –ring. Previously proposed models (Suffczynski et al. 2004; Breakspear et al. 2006) that can achieve this use strong noise in§uences that allow the system to change spontaneously between attractors corresponding to seizure-free and seizure-like activity. The stochastic nature of these transitions leads to exponentially distributed seizure durations and interseizure intervals. Although this may be applicable to absence seizures, it lacks a concrete mechanism for seizure termination and possibly cannot account for dynamics observed with focal seizures.
