小波统计方法提取想象运动诱发脑电特征

Extracting EEG Feature Evoked by Imaginary Motor Based on Wavelet Statistical Method

针对想象单侧肢体运动会导致同侧和对侧运动皮层脑区EEG信号分别出现事件相关同步(ERS)和事件相关去同步(ERD)现象,提出了一种小波变换结合统计分析的左右手运动想象诱发脑电特征提取方法.将采集的多导脑电信号进行预处理;利用db5小波,分别对左右手运动想象C3、C4导联的脑电信号能量进行6尺度小波分解,并按所需频段重构;计算不同尺度小波分解系数能量的均值、方差和相关系数,将这三个特征能量对应相加,并对C3、C4导联的能量做差,得到左右手运动想象脑电的特征向量.结果表明,该方法提取的特征向量可以较好地反映运动想象脑电事件相关同步和去同步的特点.

When the brain imagines the unilateral limb movement, EEG signal can appear event-related syn- chronization （ERS） and event-related desynchronization （ERD） in ipsilateral and contralateral motor cortex area. Based on wavelet transform combined with statistical analysis, we present a method for featuring the extraction of EEG signal evoked by left and right hand imagery motors. First, the collected multi-lead EEG signal was pretreated. Then, EEG energy signals of C3 and C4 leads were decomposed by the ＇db5＇ wavelet in the different scales. And the wavelet coefficients were reconstructed according to the desired frequency bands. Last, we calculated the mean, the variance and the correlation coefficient of different wavelet coefficients en- ergy. And these three feature energy summed correspondingly. The energy difference between the C3 and C4 lead became the feature vector of evoked EEG signal on the imaginary motor to left and right hand. The results show that this method can perform effectively to extract EEG feature and reflect ERS and ERD characteristics of EEG signal by imaginary motor.