EMD融合WPT和CSP的脑电特征提取方法  

EEG Feature Extraction of EMD Fusing WPT and CSP

作　　者：张学军 刘定宇[1] 霍延 ZHANG Xue-jun;LIU Ding-yu;HUO Yan(School of Electronics and Optical Engineering&School of Microelectronics,Nanjing University of Posts and Telecommunications,Nanjing 210023,China;Nation-Local Joint Project Engineering Lab of RF Integration&Micropackage,Nanjing University of Posts and Telecommunications,Nanjing 210023,China)

机构地区：[1]南京邮电大学电子与光学工程学院、微电子学院,江苏南京210023 [2]南京邮电大学射频集成与微组装技术国家地方联合工程实验室,江苏南京210023

出　　处：《计算机技术与发展》2020年第11期136-141,共6页Computer Technology and Development

基　　金：国家自然科学基金(61271334)。

摘　　要：脑-机接口(brain-computer interface,BCI)技术在近几十年取得了极大发展,尤其在运动障碍患者的康复训练中得到了大量的应用。脑-机接口技术包含信号采集、预处理、特征提取、分类以及外部设备控制几个部分。其中,如何更好地对信号进行特征提取并准确分类一直都是人们重点关注的问题。该文提出了一种新的特征提取算法分析运动想象(motor imagery,MI)产生的脑电波(electroencephalogram,EEG)信号,主要基于经验模式分解(empirical mode decomposition,EMD),并结合小波包变换(wavelet packet transform,WPT)和公共空间模式(common spatial pattern,CSP)。首先利用WPT将EEG信号分解为一组窄带信号并通过EMD得到相关的固有模态函数(intrinsic mode functions,IMFs),然后对每个窄带信号的IMF进行筛选,再运用CSP滤波器进行滤波获取特征,最后使用支持向量机(support vector machine,SVM)进行分类。实验应用该方法对9名受试者的运动想象脑电信号进行分类,平均准确率达95.9%,证明了该方法的可行性和有效性。Brain computer interface(BCI)technology has made great progress in recent decades,especially in the rehabilitation training of patients with sports disorders.BCI technology includes signal acquisition,signal preprocessing,feature extraction,classification and external equipment control.How to get better feature extraction and accurate classification has always been the focus of this area.We propose a new feature extraction algorithm to analyze the electroencephalogram(EEG)signals generated by motor imaging(MI)mainly based on wavelet packet transform(WPT),empirical mode decomposition(EMD)and common spatial pattern(CSP).Firstly,EEG signals are decomposed into a series of narrow band signals by WPT,and then the sub-band signals are decomposed into a set of stationary time series called intrinsic mode functions(IMFs).Secondly,appropriate IMFs are selected for signal reconstruction,and next mapped to high-dimensional space through CSP method.Corresponding feature vectors are obtained.Finally,a support vector machine(SVM)classifier is introduced in the classification experiments.The average classification accuracy of all 9 subjects is 95.9%in the experiments,which proves that the proposed method is feasible and effective.

关 键 词：脑-机接口 运动想象 经验模式分解 小波包变换 公共空间模式 

分 类 号：R318[医药卫生—生物医学工程] TP274[医药卫生—基础医学]