http://www.cs.helsinki.fi/u/ahyvarin/papers/bookfinal_ICA.pdf
Dear Sangeeta Batha,
ICA is a clever algorithm able to separate signals coming from several mixing sources, given that their physical (i.e., statistical) mean features are as less comparable as possible. A typical application of ICA approach is that of dividing each musical instrument in an orchestra when they are playing all together. First of all, the number of recording microphones must be at least not less than the number of instruments. But second of all, and this is an important condition, the instruments must perform different musical tasks; so it is practically impossible to separate two violins that are playing together exactly the same song. Concerning possible applications to EEG, the source separation ability of ICA algorithm to detect different cortical sources can providse misliding results: it is is indisputable the similarity among the EEG sources. Nevertheless, I would say: why do not try? Let me know some of your first results (at my E-mail address: [email protected] ) and good luck.
ICA is used to separate a mixture of signals in to independent sources based on statistical method. This method is also called blind source separation technique. EEG is a mixture of number of activity like eye blink potential, power line interference, neural activity etc. So we go for ica to separate the sources. The source separation is done based on statistical independence. The independence is verified by various factor like entropy.
EEGLAB is an opensource available for ICA. Also go through the ebook available online on ica for better understanding.
- Badges
- Science topic
- Similar topics
- Biological Science
- Physiology
- Biosignals
- Biomedical Signal Processing