# zbMATH — the first resource for mathematics

Phase entrainment and predictability of epileptic seizures. (English) Zbl 1026.92027
Pardalos, Panos M. (ed.) et al., Biocomputing. Conference, Univ. of Florida, Gainesville, FL, USA, February 25-27, 2001. Dordrecht: Kluwer Academic Publishers. 59-84 (2002).
Summary: Epilepsy is one of the most common disorders of the nervous system, second only to strokes. We have shown in the past that progressive entrainment between an epileptogenic focus and normal brain areas results in transitions of the brain from chaotic to less chaotic spatiotemporal states, the well-known epileptic seizures. The entrainment between two brain sites can be quantified by the T-index between measures of chaos (e.g., Lyapunov exponents) estimated from the brain electrical activity (EEG) at these sites. Recently, by applying optimization theory, and in particular quadratic zero-one programming, selecting the most entrained brain sites 10 minutes before seizures and subsequently tracing their entrainment backward in time over at most 2 hours, we have shown that over $$90\%$$ of the seizures in five patients with multiple seizures were predictable.
In this communication we show that the above procedure, applied to measures of angular frequency in the state space (average rate of phase change of state) estimated from EEG data per recording brain site over time in one of our patients with 24 recorded seizures, produces very similar results about the predictability of the epileptic seizures $$(87.5\%)$$. This finding implies an interrelation of the phase and chaos entrainment in the epileptic brain and may be used to refine procedures tot long-term prediction of epileptic seizures as well as to generate a model of the disorder within the framework of dynamical nonlinear systems.
For the entire collection see [Zbl 0995.00022].

##### MSC:
 92C50 Medical applications (general) 92C20 Neural biology 92C55 Biomedical imaging and signal processing 37N25 Dynamical systems in biology