Separating the different components of spontaneous EEG by optimized ICA

Mental EEG signal is contained by the background artifacts, the basic components of EEG rhythm and the components related to mental tasks. Many methods have been proposed to remove artifacts from EEG recordings, but there is limited when the artifacts are mixed and have comparable amplitudes with EEG, and is no helpful in separating the different components of EEG itself. Here we apply an optimized independent component analysis(ICA) method to separate these components with few channels of EEG. The result shows the separating performance is well, with components clustered and skull-projected, we also find some components related to mental tasks and their distribution. The research has some values in the cognition of mental activity and the function of brain.