睡眠脑电信号的动力学分析

对睡眠脑电信号进行了非线性动力学分析。分析中采用关联维数和Lyapunov指数的相关算法计算；出了不同睡眠期的Lyapunov指数和关联维数，讨论并分析了它们的变化趋势，总结了它们的变化规律和特点：不同的睡眠阶段很明显的对应了不同的关联维数，清醒时关联维数最大，随着睡眠的加深，关联维数降低，REM期的关联维数介于S1期和S2期之间。实验结果表明，EEG关联维数可以有效地区分大脑不同状态的特征，关联维数可以作为脑信息处理的非线性特征参数。

Analysis of the Nonlinear Dynamics for Sleep EEG

The study of the nonlinear dynamics analysis of sleep EEG is analyzed in this paper,we calculated the largest Lyapunov exponents and correlation dimension of EEG in the different sleep stages,and discussed their trends,then come to a conclusion that the sleep stages correspond to significantly different correlation dimensions.The dimensionality is highest during stage wake,and the deeper the sleep,the lower the correlation dimension.The dimensionality of sleep stage REM is between stage1 and stage2.The experiments show that the correlation dimension of EEG signal can distinguish the characteristic of brain condition effectively.The correlation dimension can be used as the nonlinear parameter of brain information processing.