脑电信号的小波变换和样本熵特征提取方法

EEG feature extraction method based on wavelet transform and sample entropy

针对现有的采用单一的特征提取算法对运动想象脑电信号识别率不高的问题,提出一种结合小波变换和样本熵的特征提取方法.通过小波变换,把脑电信号进行3层分解,抽取出对应于脑电β节律频带的小波系数的能量均值和能量均值差,并结合脑电信号的样本熵组成特征向量,使用支持向量机分类器对左右手运动想象脑电信号进行分类.结果表明,结合小波变换和样本熵的特征提取方法明显优于仅采用小波变换、样本熵以及其他传统的特征提取方法,得到的最高正确识别率为91.43%.

Considering the issue of low recognition rate for electroencephalograph（EEG） signal of motor imagery by using current single feature extraction method,a feature extraction method based on wavelet transform and sample entropy is presented in this paper.The EEG signals are decomposed to three levels by wavelet transform and the average energy and its difference of wavelet coefficient corresponding to the β rhythm of EEG signals are computed.The feature vector is composed of the average energy,its difference of wavelet coefficient and sample entropy of EEG signals.Finally,the left-right hands motor imagery EEG signals are classified by a support vector machine classifier.The experimental results show that the feature extraction method combining wavelet transform and sample entropy is much better than the ways of only using wavelet transform,sample entropy,or others,and its highest recognition rate is 91.43%.