Modeling and analyzing non-seizure EEG data for patients with epilepsy

摘要

We present nonlinear analysis of non-seizure electroencephalogram (EEG) time series data from four epileptic patients. A non-seizure state is a period that is free of any part of an epileptic seizure, including the transition to a fully developed episode. EEG measurements are typically contaminated with a large amount of non- neurophysiological source information, generally called artifact, which arises, for example, from eye movement, muscle tension, and physical motion. The first objective of this study is to gain some insight into how much variability in analysis results to be expected from patients having similar clinical characteristics. The second objective is to investigate the impact of eye movement on the analysis results. A special feature presented here is the introduction and testing of a filter for eye movement artifact. The third objective is to determine if neurophysiological activity as viewed from two adjacent channels appears dynamically to be the same.