Suitable ICA algorithm for extracting saccade-related EEG signals

Arao Funase, Motoaki Mouri, Andrzej Cichocki, Ichi Takumi

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)

Abstract

Our goal is to develop a novel BCI based on an eye movements system employing EEG signals on-line. Most of the analysis on EEG signals has been performed using ensemble averaging approaches. However,It is suitable to analyze raw EEG signals in signal processing methods for BCI. In order to process raw EEG signals, we used independent component analysis(ICA). However, we do not know which ICA algorithms have good performance. It is important to check which ICA algorithms have good performance to develop BCIs. Previous paper presented extraction rate of saccade-related EEG signals by five ICA algorithms and eight window size. However, three ICA algorithms, the FastICA, the NG-FICA and the JADE algorithms, are based on 4th order statistic and AMUSE algorithm has an improved algorithm named SOBI.Therefore, we must re-select ICA algorithms. In this paper, we add new algorithms; the SOBI and the MILCA. The SOBI is an improved algorithm based on the AMUSE and uses at least two covariance matrices at different time steps. The MILCA use the independency based on mutual information. Using the Fast ICA, the JADE, the AMUSE, the SOBI, and the MILCA, we extract saccade-related EEG signals and check extracting rates. Secondly, in order to get more robustness against EOG noise, we use improved FastICA with reference signals and check extracting rates.

Original languageEnglish
Title of host publicationNeural Information Processing - 16th International Conference, ICONIP 2009, Proceedings
Pages409-416
Number of pages8
EditionPART 1
DOIs
Publication statusPublished - 2009
Externally publishedYes
Event16th International Conference on Neural Information Processing, ICONIP 2009 - Bangkok, Thailand
Duration: 1 Dec 20095 Dec 2009

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 1
Volume5863 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference16th International Conference on Neural Information Processing, ICONIP 2009
Country/TerritoryThailand
CityBangkok
Period1/12/095/12/09

Fingerprint

Dive into the research topics of 'Suitable ICA algorithm for extracting saccade-related EEG signals'. Together they form a unique fingerprint.

Cite this