An Intelligent Brain Computer Interface of Visual Evoked Potential EEG

The goal of our research is to design an EEG BCI (brain computer interface) system and develop techniques for helping the serious disabled with spine/central nerve injury, motor neuron disease or without upper limb/foot. In our research, the brain wave amplifier we used is the NuAmps amplifier and the acquisitive/analytic software for EEG is named "Scan 4.3". Both of them were developed by NeuroScan Company. The other software we used to implement the human-machine interface and on-line analysis functions is LabVIEW 8.2.1 version developed by National Instruments. We programmed TCP/IP communication protocol to connect Scan 4.3 with LabVIEW. In the experiments, we focused on VEP (visual evoked potential) to design the EEG BCI system. There are four directional arrows (up, down, left, right) shown on the monitor randomly for stimulating the subject. Regarding data analysis, we used ICA (independent component analysis) to filter the artifact of EOG/EMG, and calculate the average value after the summation of all trials of EEG signals. Therefore, we found some special phenomena and regarded them as VEP EEG features. We applied these VEP EEG features to design a directional examination for 5 male subjects. The average of recognition ratio of EEG resulted from VEP is over 90 % if the subject can be concentrative during the test.