摘要
脑电信号极易受到眼电信号的干扰,这会导致脑电信号处理结果与实际情况发生较大的偏差,因此,去除包含于脑电信号中的眼电成分是信号预处理的一个重要操作。研究了独立成分分析理论及概要模型,提出一种基于Informax的优化ICA方法,对混入脑电信号中的眼电信号进行辨别、分离、重构,实验结果表明该方法能够准确地从混合信号中区分出眼电伪迹的独立成分,进而实现对原脑电信号的特征增强。
The recorded Electroencephalography(EEG)signals are easily contaminated by ocular artifacts,which results in a large deviation between the results of processing EEG signal and the actual situation.Therefore,the removal of electrooculogram artifacts components contained in EEG signals is an important operation during signal preprocessing.In this paper,the theory and the principle model of independent component analysis(ICA)are studied.An optimization ICA method based on Informax is proposed to identify,separate and reconstruct electrooculogram artifacts which mixed with EEG signals.The experimental results show that this method can accurately distinguish the independent components of electrooculogram artifacts from the mixed EEG signals,so as to realize the enhanced features of the original EEG signals.
作者
耿晓中
李得志
GENG Xiao-zhong(School of Computer Technology and Engineering,Changchun Institute of Technology,Changchun 130012,China)
出处
《长春工程学院学报(自然科学版)》
2020年第1期78-81,共4页
Journal of Changchun Institute of Technology:Natural Sciences Edition
基金
吉林省科技厅项目(20190302110GX)。
关键词
脑电信号
眼电伪迹
独立成分分析
信息极大化
electroencephalography signal
electrooculogram artifact
independent component analysis
information-maximization