基于样本熵的睡眠脑电分期

发布时间：2018-07-02 08:17

  本文选题：睡眠分期 + 样本熵　； 参考：《江苏大学学报(自然科学版)》2009年05期

【摘要】：运用样本熵从波士顿Beth Israel睡眠脑电实验数据中提取睡眠特征值,对睡眠分期进行研究.针对脑电属于微弱非平稳随机信号、难于提取特征的特点,利用小波变换先有效地消除脑电信号中的噪声,再计算其样本熵用以表征睡眠各分期.计算结果表明,由清醒期到非快速眼动的Ⅳ期过程中,其样本熵值呈规律性逐渐变小,与该库中专家评定的结果相符.这说明经过小波消噪和样本熵处理的脑电信号能准确地反映睡眠各期的变化特征,比用近似熵表征睡眠分期更准确、运算速度更快,完全适用于非平稳随机信号的处理.
[Abstract]:The sleep stages were studied by extracting the sleep characteristic values from the data of Beth Israel sleep EEG experiment in Boston using sample entropy. Because EEG is a weak non-stationary random signal, it is difficult to extract features. Wavelet transform is used to effectively eliminate the noise in EEG signal, and then the sample entropy is calculated to characterize the sleep stages. The calculated results show that the entropy values of the samples decrease gradually during the period 鈪，

本文编号：2089500