摘要
眼电(EOG)伪迹是基于脑电(EEG)的脑机接口系统中最重要的干扰。为自动去除这种干扰,提出了基于ICA-RLS的EOG伪迹自动去除算法。首先,ICA将多个导联EEG信号分解成数目相等的若干个独立分量(ICs)。然后,计算并依据每个IC的峰态系数,自动地从ICs中识别出EOG分量。最后,识别出的EOG分量用作RLS自适应滤波器的参考信号,并用该滤波器对原始EEG进行滤波,在无需记录EOG情况下实现EOG伪迹的自动去除。用提出算法对2008年脑机接口竞赛数据进行处理,从去除伪迹后信号的观察、信息保留完整性和最终分类结果的计算进行评价。与标准的ICA算法相比,提出算法能够更好去除EOG伪迹,同时获取更高的分类正确率。
Electrooculogram（ EOG） artifact is the most important form of interferences in electroencephalogram（ EEG） based brain computer interfaces（ BCIs）.In order to automatically remove the EOG artifact interference,an EOG artifact automatic removal algorithm based on independent component analysis-recursive least squares（ ICARLS） is proposed.Firstly,ICA is used to decompose the multiple channel EEG signals into an equal number of independent components（ ICs）.Then,the kurtosis value of each IC is calculated,based on which the EOG artifacts can be identified from the ICs.Finally,the identified EOG artifacts are used as the reference signals of the adaptive filter based on RLS; and the filter is used to filter the original EEG signal to reduce interference.After filtering,the EOG artifacts are automatically removed without the need of recording EOG.The EEG data from 2008 BCI competition were processed using the proposed algorithm and evaluated from observing the signal after removing the EOG artifact,computing the integrity of the signal and classification result.Compared with standard ICA algorithm,the proposed algorithm can better remove EOG artifacts and achieve higher classification accuracy.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2015年第3期668-674,共7页
Chinese Journal of Scientific Instrument
基金
国家自然科学基金(60975079
31100709)
上海市浦江人才计划(14PJ1431300)资助项目
