Figure 5.

Enhancement of EEG via BSS. First, the raw EEG data (sensor signals) is decomposed and ranked as independent or spatially decorrelated components; in the next step, only the useful components are projected back to the scalp level, while undesirable components containing artifacts and noise are removed from the signal. The main advantage of our approach is that we do not need any expert decision to select significant components, since the AMUSE algorithm automatically ranks the components as illustrated in Figure 6.