Feature extraction of evoked potential using infomax algorithm of independent component analysis
Liyu Huang · Journal of Northwest University · 2009
Aim To present an infomax algorithm for ICA used for noise removal and signal feature extraction of Event-Related Potentials(ERP).Methods This algorithm is used to decompose electroencephalogram(EEG) data in order to remove the artifacts and select main components related to Visual Evoked Potential(VEP).Results The selected components are projected to the electrodes on the scalp.Conclusion The ICA algorithm can effectively extract and improve the ERP signal features,and enhance the exact orientation of the P3 element.