小波分析和独立分量分析结合的诱发电位信号提取研究  

Extraction of Evoked Related Potentials by using the Combination of Independent Component Analysis and Wavelet Analysis

作　　者：邹凌[1,2] 陈树越[1] 孙玉强[1] 马正华[1] 

机构地区：[1]江苏工业学院信息学院,常州213164 [2]北京师范大学认知神经科学与学习国家重点实验室,北京100875

出　　处：《生物医学工程学杂志》2010年第4期741-745,共5页Journal of Biomedical Engineering

基　　金：北京师范大学认知神经科学与学习国家重点实验室开放课题资助;北京师范大学应用实验心理北京市重点实验室开放研究基金课题资助;江苏省青蓝工程资助

摘　　要：提出将独立分量分析(ICA)和小波分析结合的方法提取诱发电位(ERPs)信号。首先采用扩展信息最大化ICA算法处理脑电信号,得到独立成分;然后将小波阀值收缩算法作为中间步骤处理独立成分;对新的独立成分进行ICA的逆变换以重建脑电数据;少次叠加消噪后的脑电数据,提取出ERPs信号。真实数据的处理结果表明:ICA算法和结合方法都可以有效去除混合在ERPs信号中的眼电噪声和肌电噪声,但结合算法可以更好地保留噪声独立成分中的脑神经活动,实现微弱ERPs信号的有效提取。In this paper we present a new method of combining Independent Component Analysis（ICA） and Wavelet de-noising algorithm to extract Evoked Related Potentials（ERPs）.First,the extended Infomax-ICA algorithm is used to analyze EEG signals and obtain the independent components（Ics）;Then,the Wave Shrink（WS） method is applied to the demixed Ics as an intermediate step;the EEG data were rebuilt by using the inverse ICA based on the new Ics;the ERPs were extracted by using de-noised EEG data after being averaged several trials.The experimental results showed that the combined method and ICA method could remove eye artifacts and muscle artifacts mixed in the ERPs,while the combined method could retain the brain neural activity mixed in the noise Ics and could extract the weak ERPs efficiently from strong background artifacts.

关 键 词：扩展信息最大化独立分量分析 小波阀值收缩 结合方法 事件相关诱发电位 去噪 

分 类 号：R318.0[医药卫生—生物医学工程] TN911.7[医药卫生—基础医学]