EEG feature detection and classification algorithm in brain-computation interface

Brain-computer interface research focused on using electroencephalogram(EEG) from the scalp over sensorimotor cortex to control outer device. The studies seek to improve the classification accuracy by improving the selection of signal features based on non-linear methods. Since EEG signals may be considered chaotic, chaos theory may supply effective quantitative descriptors of EEG dynamics and of underlying chaos in the brain. The complexity of the chaotic system can be characterized by complexity measure computed from the signals generated by the system.Two new features of EEG, Kolmogorov and CO complexity measure are presented for analyzing EEG signals in BCI system. The experiments proved that the method is effective; the accuracy of the system reaches 90.3%.