EEG COMPLEXITY TOPOGRAPHY
复杂度脑电地形图研究

EEG topography is one of the hotspots in EEG analysis. Through analysis and comparison of several kinds of complexity measure algorithm, we found that approximate entropy needs shorter time series so that some difficulty owing to the nonstationarity could be overcome; in addition, no coarse graining preprocessing is needed. Therefore it has some virtues in complexity analysis of biomedical signals. First we compute the complexity measures for several channels of EEG signals, then through interpolation, construct dynamic complexity topography so as to observe relative intensities among different parts of EEG signal complexity at the same time and their changes with time. Through analysis of some patients' EEG data, we explored possible difference between abnormal and normal subjects in complexity topography, from which some information for diagnosing brain disease especially for some functional disease was extrected. We found that topography pattern of schizophrenia patient with eye closed was more complex than that of normal subjects. We also found the complexity level would decrease during epileptic seizure capture.