摘要
为了有效获取脑电信息,提取多模式类(左手,右手,舌,脚)想象运动诱发脑电特征,提出了一种小波-共空间模式分析的四类运动想象诱发脑电特征提取方法.首先,使用db4小波对预处理后纯净的脑电信号进行5层小波分解,获得小波变换的各层逼近系数和细节系数,去除冗余频段,并进行重构.然后,采用"一对多"共空间模式方法,构建多个空间滤波器,对滤波信号进行共空间模式映射,得到每种想象运动模式下的投影信号.最后,对该投影信号能量求方差取对数,组成特征值向量,再进行差值处理获得多模式类运动想象脑电特征.仿真实验结果表明,该方法可以较好的提取多类运动想象脑电信号特征.
In order to acquire the effective EEG information and extract features of EEG signal evoked by multi classification（left hand,right hand,tongue and feet）imagery motor,this paper proposed a method to wavelet combined with common spatial pattern analysis for the feature extraction of EEG evoked four kinds of imaginary motion.First of all,apply db4 wavelet by 5layer wavelet decomposition to pre-processing EEG signal,which obtain the wavelet approximation coefficients and detail coefficients,remove of redundant frequency bands and reconstruct EEG signals.Then,using the＂one-versusrest＂method of common spatial pattern（CSP）,we constructed multiple CSP filters,get projection signal of each kind of motor imagery modes by CSP mapping.Finally,obtain feature vectors by calculating variance to the projection signal energy and doing logarithm operation.After the difference processing to the vector,EEG features for different motor imagery modes were extracted.Simulation results showed that this method can effectively perform feature extraction for Multi-mode motor imagery extract signal characteristics of multiple motor imagery.
出处
《测试技术学报》
2015年第6期523-528,共6页
Journal of Test and Measurement Technology
基金
国家基础科学人才培养基金资助项目(J1103210)
山西省自然科学基金资助项目(2013011016-2)
关键词
运动想象
小波变换
共空间模式
特征提取
诱发脑电
motor imagery
wavelet transform
common spatial pattern
feature extraction
evoked potentials
