摘要
针对癫痫脑电(EEG)信号的非平稳性和非线性,提出一种基于集合经验模式分解(EEMD)脑电的方法,首先利用EEMD将EEG信号分解,得到各阶本征模式分量(IMF),然后提取有效特征,构成特征分量,最后用支持向量机(LS-SVM)对其分类;采用德国波恩癫痫研究室临床采集的癫痫脑电数据库,实验结果表明:特征提取方法对癫痫发作间歇期和发作期EEG的分类正确率最高可达99.5%。
According to non-stationary and nonlinear feature of epileptic EEG signals,this paper proposes a kind of EEG method based on ensemble empirical mode decomposition( EEMD) method,firstly uses EEMD to decompose EEG signal to obtain intrinsic mode function of each period,then extracts effective feature to constitute feature function,and finally uses support vector machine to make classification based on clinical epileptic EEG data collected by German Bonn Epilepsy Research Laboratory. Experiment results show that this feature extraction method has 99. 5 percent of accuracy for the classification of EEG at intermittent period and paroxysm period of epileptic paroxysm.
出处
《重庆工商大学学报(自然科学版)》
2014年第5期90-94,共5页
Journal of Chongqing Technology and Business University:Natural Science Edition
基金
淮南师范学院校级项目(2011LK93q)
关键词
癫痫脑电信号
集合经验模式分解
最小二乘支持向量机
本征模式分量
epileptic EEG signal
ensemble empirical mode decomposition(EEMD)
least square support vector machine
intrinsic mode function
作者简介
李营(1983-),女,陕西省铜川市人,助教,硕士,从事信号与信息处理研究.