Online BCI with Stable Sources

In this paper, we show that the estimated intra-cranial sources using source localization on EEG signals can be used for online Brain Computer Interation (BCI) and the discriminant sources remain stable over days. Classifiers are trained on discriminant sources obtained for Error-related Potential (ErrP) based BCI on day 1 and then tested online on day 2. The results for nine subjects show that the source localization with discriminant sources can be used for online detection of ErrPs. Furthermore, the number of common sources among the discriminant sources for different days is above chance which shows that the discriminant sources remain stable over days.