Research of Brain-Computer Interface Based on the Time-Frequency-Spatial Filter

In order to improve the accuracy of the brain state classification, the time-frequency-spatial filter algorithm is put forward. In this algorithm, the signal features are extracted in terms of time, frequency and space. The parameters of the spatial pattern filter are not fixed, but variable with time. Specific frequency bands are also optimized simultaneously in this algorithm. So it can perfect the BCI pattern. It is applied to BCI datasets to illustrate the performance and validity of the algorithm. And the results indicate the algorithm can improve the accuracy of classification. The method will facilitate to classify EEG signals with small training sets.