The analysis of “yes” and “no” response by visual evoked stimulations in human EEG

Many diseases can damage the brain, which make brain unable to communicate normally with the outside world through the peripheral nervous system. As an emerging technology, brain-computer interface(BCI) can convert electroencephalogram(EEG) to the control signal to try repairing function for patients. So the study of BCI can improve the life quality of the patients with severe neuromuscular damage. This paper introduces the EEG data acquisition system capable of detecting the visual evoked EEG of an answer to “yes/no” question. EEG signal processing includes preprocessing, common spatial pattern (CSP)-based feature extraction algorithm and support vector machine (SVM)-based feature classification method. Finally classification accuracy can achieves 81% in the experimental data of healthy subjects. The experimental results show that the EEG analysis method makes the patients with neuromuscular system damage simply communicate with external world, and provides the basis for study of further BCI.