A Modified Fast Independent Component Analysis and its Application to ERP Extraction

Feature extraction of event related potential (ERP) plays an important role in both fundamental and clinical research for cerebral neurophysiology. ICA is a method for separating blind signals based on signal statistic characteristics. In this paper, the fundamental, discrimination condition and pra...

Full description

Saved in:
Bibliographic Details
Published in2009 2nd International Conference on Biomedical Engineering and Informatics pp. 1 - 4
Main Authors Binfeng Xu, Yarong Hu, Xiaogang Luo, Chenglin Peng
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.10.2009
Subjects
Brain modeling
Convergence
Educational institutions
Electroencephalography
Enterprise resource planning
Independent component analysis
Iterative algorithms
Kernel
Scalp
Signal processing algorithms
Online AccessGet full text

Cover

More Information
Summary:Feature extraction of event related potential (ERP) plays an important role in both fundamental and clinical research for cerebral neurophysiology. ICA is a method for separating blind signals based on signal statistic characteristics. In this paper, the fundamental, discrimination condition and practical algorithm of Independent Component Analysis are discussed. Then, a fast Independent Component Analysis algorithm (FastICA) is introduced. But like FastICA, its convergence is dependent in initial weight. By importing loose gene in the algorithm, the new algorithm could implement convergence in large-scale. By modifying kernel iterate course, several iterations of FastICA are merged into one iteration of Modified FastICA, so the convergence of ICA will be accelerated. Finally, Modified ICA is applied to ERP extraction. The simulation shows that using the improved algorithm convergence speed can be increased.
ISBN:9781424441327
1424441323
ISSN:1948-2914
1948-2922
DOI:10.1109/BMEI.2009.5305610