EEG Signal Processing And The Research On BCI System Based On ALPHA Wave

Brain computer interface(BCI) is a new information exchange technology of direct exchange of information and control path based on human brain and a computer or other electronic devices without the normal brain output channels (peripheral nerve and muscle tissue). The BCI system is usually composed of signal acquisition, signal processing (preprocessing, feature extraction, classification algorithm) and the command output units. Therefore, brain-computer interface technology involves multiple disciplines, including biomedical engineering, pattern recognition, computer technology, and communication technology. Currently, the main purpose of researching the brain computer interface is to provide a special way to exchange information for the people who lose the capability of moving but can think as normal. BCI system can also provide an additional means of controlling a game for people to provide a new form of entertainment.Signal processing is important, because it has a great influence on system performance. Preprocessing includes amplification, filtering and artifact elimination. Feature extraction is to extract of the particular EEG produced by human brain, but not to interpret the EEG itself. Classification algorithm is to classify the extracted features in order to generate the command signal.We use the spontaneous EEG ofαwave which has a high amplitude, and another advantage ofαwave is that it do not need external incentives. Theαwave has a special blocking property, when the eyes open and eyes closed there will be significant changes in the amplitude of the signal, we do the feature extraction just based on this property. We eliminate the artifact by using fast independent component analysis (FastICA), and then filter the signal by using the wavelet transform, at last we do the feature extraction by using power spectrum estimation method. We have made the relevant experimental software and hardware platform. To improve the system performance, we use multi-level amplification, band pass filter and double-T notch circuit in signal processing platform. In order to improve the accuracy and anti-jamming capability of the signal acquisition system, we use high-precision 24-bit A/D converter ADS1256, which is controlled by the low power MSP430 microcontroller through the SPI bus for control and data transmission, then the data is transferred to the computer through the RS232 bus.The system can simultaneously display the being processed images of the EEG and the corresponding state of feature extraction. It has high anti-interference ability and recognition rate.