An adaptive filter bank for motor imagery based Brain Computer Interface

Brain Computer Interface (BCI) provides an alternative communication and control method for people with severe motor disabilities. Motor imagery patterns are widely used in Electroencephalogram (EEG) based BCIs. These motor imagery activities are associated with variation in alpha and beta band power of EEG signals called Event Related Desynchronization/synchronization (ERD/ERS). The dominant frequency bands are subject-specific and therefore performance of motor imagery based BCIs are sensitive to both temporal filtering and spatial filtering. As the optimum filter is strongly subject-dependent, we propose a method that selects the subject-specific discriminative frequency components using time-frequency plots of Fisher ratio of two-class motor imagery patterns. We also propose a low complexity adaptive Finite Impulse Response (FIR) filter bank system based on coefficient decimation technique which can realize the subject-specific bandpass filters adaptively depending on the information of Fisher ratio map. Features are extracted only from the selected frequency components. The proposed adaptive filter bank based system offers average classification accuracy of about 90%, which is slightly better than the existing fixed filter bank system.

[1]  E Donchin,et al.  Brain-computer interface technology: a review of the first international meeting. , 2000, IEEE transactions on rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society.

[2]  Dennis J. McFarland,et al.  Brain-computer interface (BCI) operation: signal and noise during early training sessions , 2005, Clinical Neurophysiology.

[3]  G. Pfurtscheller,et al.  Optimal spatial filtering of single trial EEG during imagined hand movement. , 2000, IEEE transactions on rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society.

[4]  Cuntai Guan,et al.  Filter Bank Common Spatial Pattern (FBCSP) in Brain-Computer Interface , 2008, 2008 IEEE International Joint Conference on Neural Networks (IEEE World Congress on Computational Intelligence).

[5]  Klaus-Robert Müller,et al.  Spatio-spectral filters for improving the classification of single trial EEG , 2005, IEEE Transactions on Biomedical Engineering.

[6]  Rodica Strungaru,et al.  EEG Features Extraction for Motor Imagery , 2006, 2006 International Conference of the IEEE Engineering in Medicine and Biology Society.

[7]  G. Pfurtscheller,et al.  Brain-Computer Interfaces for Communication and Control. , 2011, Communications of the ACM.

[8]  Ping Xue,et al.  Sub-band Common Spatial Pattern (SBCSP) for Brain-Computer Interface , 2007, 2007 3rd International IEEE/EMBS Conference on Neural Engineering.

[9]  G. Pfurtscheller,et al.  Motor imagery activates primary sensorimotor area in humans , 1997, Neuroscience Letters.

[10]  Klaus-Robert Müller,et al.  Combined Optimization of Spatial and Temporal Filters for Improving Brain-Computer Interfacing , 2006, IEEE Transactions on Biomedical Engineering.

[11]  A. Prasad Vinod,et al.  Coefficient decimation approach for realizing reconfigurable finite impulse response filters , 2008, 2008 IEEE International Symposium on Circuits and Systems.