A NEW METHOD OF PHASE SAPCE ANALYSIS OF EEG SIGNAL
- 物理技术－已发表论文 
提出一种新的脑电(EEG)信号相空间分析方法。通过计算相空间状态点间的欧氏距离,定义相对嵌入维数的态密度和态方差,并与反映非线性动力学系统混沌特征的关联维数作比较。对各种实测的脑电时间序列的计算结果表明,态密度和态方差不仅计算简单,计算结果一致可靠,而且和关联维数相比,更能有效地反映非线性动力学系统的某些特征。此外,计算了基于距离协方差的脑电信号的奇异谱,并对实验结果作了分析。A new method of phase space analysis of EEG is proposed. By calculating the Euclidian dis-tances of the phase point in phase space, both the phase density and the phase variance are defined and compared with correlation dimension which reflects the chaotic features of nonlinear dynamics system. By calculating various real time series of EEG, the results show that the phase density and phase variance are not only simple in calculation, but also more effective in reflecting the features of nonlinear dynamics system than the correlation dimension. Furthermore, the singular spectra for both the normal EEG and the epileptic EEG are calculated and illustrated based on the covariance of distance of phase space, and the analysis of the results is given.