Application of Support Vector Machine for the classification of sensorimotor rhythms in Brain Computer Interface

An approach of Brain Computer Interface (BCI) is to replace the motor function. The motor imagery itself is a common strategy for motor rehabilitation. The aims of this paper are the extraction and classification of sensorimotor rhythms, mu rhythm (8-12 Hz) and beta rhythm (12-30 Hz), which occur during motor imagery of the hands. To achieve these purposes, four steps were undertaken: data acquisition, pre-processing, extracting of the features and classification. The Support Vector Machine (SVM) classifier was applied on the extracted features after using multiresolution wavelet analysis. Thus, the higher rate of classification is 86%, and the average for the 25 subjects is 79%, 10% higher than the average for the classifier within BCI2000 software.