Design Of The Brain-Computer Interface System Based On USB Technique

Brain-Computer Interface (BCI) is a direct information communication andcontrol channel established between human and computer or other electronics devicesand it is a wholly new communication system that does not depend on the brain'snormal output pathways of peripheral nerves and muscles. As a wholly newinformation communication and control technology, BCI is winning more and moreattention. EEG-based BCI may not only provide the paralyzed with an effectivecommunication and control channels with outside world, but also can be used in theother fields such as spaceflight (for example: an airplane is controlled directed byhuman thoughts).In terms of the fact that event-related potential(P300)which reflects psychologyof human beings will come out by some optical stimulation , this dissertation presentsa simple and easy P300 based BCI system combined with USB technique. While aperson are gazing at one letter on the Oddball-Stimulating-Screen the single P300wave appears, and then EEG signals are transmitted to the computer. Aftercomplicated analysis we will obtain the information that which row and whichcolumn the letter focused by the person is on. For example: the "O" letter is on thethird row and the third column. At last the signal like "33" is sent to code-system. Ifthe letters are the same to those set in advance, the lock will be opened or the letterswill be printed.In this thesis,the BCI system is expounded by dividing into two parts. One ofthem is called "software part" for signal processing, and the other is called "hardwarepart" for peripheral equipments design.In the "software part" of BCI system, the wavelet analysis and the ICA algorithmare improved by listing the examples of P300 that reflects the related potential of thebrain's recognizing rare events. As for wavelet transform, we at first remove theinterference of EEG background mixed with P300, then rebuild the de-noised data anddetect the P300 peaks. After that, with the helps of time-shift correction of P300latencies and the wavelet reconstruction, we can get a rather satisfied P300 waveformthat only can be obtained by the conventional coherent averaging needed over manytrials. As for ICA, a method based on the Extended Informax of ICA is proposed forseldom trials extraction of ERP. In this algorithm, we at first remove the interferencesuch as blink artifacts and power noise, then rebuild the EEG data and at last add theseveral trails in order to get a rather satisfied P300 wave patter which is similar to thatobtained by the conventional coherent averaging needed over 36 times.In the "hardware part" of BCI system, both the collection platform and the codesystem are designed based on USB technique. The amplificatory EEG signals arecarried to the computer for processing. Then signals are displayed and saved in theenvironment provided by VB software. After pattern recognition the new signals,which contains information of the letters chosen by people facing aOddball-stimulating-screen, are sent to peripheral code-system and displayed onLEDs. Meanwhile, we've added a keyboard for setting or changing codes at anymoment. The whole hardware system can work without the development environmentand can be communicated by Windows of the computer as a USB equipment.