Inverse Problem Applied to BCI's : Keeping Track of the EEG's Brain Dynamics Using Kalman Filtering

In this paper, we use Kalman filtering to improve inverse solutions for BCI applications. The algorithm is tested on EEG data from the BCI 2003 competition. The aim of this work is to improve the spatial resolution of the EEGs using the inverse model and take profit of their excellent time resolution with the Kalman filter. Preliminary results show a 4% improvement with our additionnal Kalman filtering