Blind source separation for fMRI signals using a new independent component analysis algorithm and principal component analysis

Weiwei Zhang, Zhenwei Shi, Huanwen Tang, Yiyuan Tang

Research output: Contribution to journalArticle

1 Scopus citations

Abstract

The application of independent component analysis (ICA) to the functional magnetic resonance imaging (fMRI) data can separate many independent sources. But in the processing there are two difficulties: (1) the data of the fMRI is usually on a large scale, so the computing is time-consuming; (2) we cannot avoid the errors for too heavy computational load, this brings many troubles. Thus we think of reducing the data. In this article we used the standard information theoretic methods to estimate the number of the sources and used the principal component analysis (PCA) to reduce the data. By this process, we estimated the number of the sources and reduced the data successfully; Then we applied the ICA algorithm to the reduced fMRI data; this method raised the speed of operation. After application of the new ICA algorithm and another algorithm (FastICA) to the fMRI data, a comparison was made. The results show that the new algorithm can separate the fMRI data fast and effectively and it is superior to the FastICA on the accuracy of estimating the temporal dynamics of activations.

Original languageEnglish
Pages (from-to)430-433
Number of pages4
JournalSheng wu yi xue gong cheng xue za zhi = Journal of biomedical engineering = Shengwu yixue gongchengxue zazhi
Volume24
Issue number2
StatePublished - Apr 2007

Fingerprint Dive into the research topics of 'Blind source separation for fMRI signals using a new independent component analysis algorithm and principal component analysis'. Together they form a unique fingerprint.

  • Cite this