计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (31): 218-220.DOI: 10.3778/j.issn.1002-8331.2009.31.065

• 工程与应用 • 上一篇    下一篇

典型相关分析去除脑电信号中眼电伪迹的研究

张 莉1,2,何传红1,2,何 为1,2   

  1. 1.重庆大学 电气工程学院,重庆 400030
    2.重庆大学 输配电装备及系统安全与新技术国家重点实验室,重庆 400030
  • 收稿日期:2009-05-07 修回日期:2009-06-12 出版日期:2009-11-01 发布日期:2009-11-01
  • 通讯作者: 张 莉

Research on removing EOG artifacts from EEG based on CCA

ZHANG Li1,2,HE Chuan-hong1,2,HE Wei1,2   

  1. 1.College of Electrical Engineering,Chongqing University,Chongqing 400030,China
    2.State Key Lab of Power Transmission Equipment & System Security and New Tech.,Chongqing University,Chongqing 400030,China
  • Received:2009-05-07 Revised:2009-06-12 Online:2009-11-01 Published:2009-11-01
  • Contact: ZHANG Li

摘要: 给出了一种基于典型相关分析(Canonical Correlation Analysis,CCA)的盲源分离技术来去除脑电信号中的眼电伪迹。通过实验验证了基于CCA的盲源分离方法去除眼电伪迹的有效性,并将该方法与广泛使用的独立分量分析(Independent Component Analysis,ICA)进行了比较。实验结果表明,基于CCA的盲源分离方法可以对眼电伪迹进行成功地分离和消除,该方法相较于ICA方法而言,算法更为简单,计算速度更快。

关键词: 脑电信号, 眼电伪迹, 典型相关分析, 盲源分离

Abstract: Canonical Correlation Analysis(CCA) as a Blind Source Separation(BSS) technique is applied to the removal of Electroencephalography(EEG) artifacts.This method is tested and compared with the widely used Independent Component Analysis(ICA) method through experiments.The experiment results show that the CCA-based method is effective in separating and eliminating Electrooculography(EOG) contamination.Compared with ICA method,the proposed method has the advantages of simplicity and high speed.

Key words: Electroencephalography(EEG), EOG artifacts, Canonical Correlation Analysis(CCA), Blind Source Separation(BSS)

中图分类号: