用相空间内邻近轨线伸展的相关性研究脑电的混沌特性

本文的研究针对目前脑电的非线性动力学研究中一个尚未定论的关键问题：脑电究竟是混沌信号还是无序的随机信号？文中介绍了作者的研究成果：用相空间中一对邻近轨线伸展的伸关性来鉴别混沌和无序的随机运动，并应用这种新的混沌行为分析方法证明了脑电的混沌特性，作为比较和验证，本文也介绍了用该方法对加上不同水平噪声和Lorenz系统仿真的研究结果。

TESTING FOR CHAOTIC DYNAMICS IN EEG VIA CORRELATION OF LOCAL STRETCHINGS

Correlation of two local stretchings (CTLS) is used as a way of characterizing chaotic dynamics in EEG. With a comparison of localized divergences of two neighboring trajectories from a fiducial trajectory, chaos is determined when the correlation is strong or the source is noise with little or no correlation. The method was demonstrated for Lorenz systems with various noisy level, and used for 100 EEG data sets from 3 groups: the first group was from an epileptic responsible focus of partial epileptic (PE) patients, the second group was away from the focus of PE patients, and the third group was from the normal EEG. The results showed that the EEGs of all cases were chaotic.