大腦人機介面是一種利用大腦直接控制外在儀器（操作手臂、控制輪椅）的一項技術，經由腦波辨識系統讓使用者可以透過思想就達到與外界溝通的目的，希望藉由這項技術能有效的幫助因中樞神經或肌肉受損而無法擁有自主行為的病人（如中樞神經系統損傷、重度中風、有思想和意識而無行為能力的病人等）直接利用知覺以及認知能力功能都健全的大腦來進行與外界溝通，而大腦人機介面就是很好的媒介。
本研究利用視覺誘發回溯系統誘發腦波，擷取誘發的腦波P300訊號再透過適應性類神經模糊推論系統(Adaptive Neuro-Fuzzy Interface System, ANFIS)做分類，分類結果可以做為大腦人機介面控制源，並以有效提升腦波P300訊號之辨識率為目標。從實驗結果得知，利用適應性類神經模糊推論系統來分類腦電波，單筆腦波平均辨識率為79%，最高正確率為83%。將二筆腦波平均後作辨識則平均辨識提高為87%，最高可達95%。然而本研究之受測者均為四肢健全腦部功能無受損者來進行實驗，且實驗資料濾除過程嚴謹，凡是受到干擾之腦波訊號皆去除。期待本篇研究能對致力於發展大腦人機介面的研究者或有興趣的學者提供幫助。

The brain computer interface is a technology that through human being’s brain to control tools ( machine、wheel chair etc.), by using a recognition system for brain, the users can communicate with others through their thoughts. The technology can be used to help the patients who can’t move by themselves but the functions of their brains are fine. They can do what they want to do depending on this technology. The brain computer interface is appropriate for this application.
Our research uses visual evoked feedback system to evoke brain wave and we extract the P300 signals as the input to the classifier Adaptive Neuro-Fuzzy Interface System, the results after classifying can be taken as the control sources of the brain computer interface. Our purpose is to improve the classifying accuracy for the brain computer interface. The experiment results show that the average accuracy of single brain wave is 79% and the best accuracy is 83%, the average of two brain waves is 87% and the best accuracy is 95%. Nevertheless, since the subjects of our research are healthy and the data we used are those without disturbance, the data of the study should be cautiously made. Our research is beneficial for people who are interested in brain computer interface.

[1]吳進安等．神經診斷學．國立編譯館． 1998。
[2]林景福等．圖解腦波入門．合記圖書出版社． 1985。
[3]B.J. Fisch. “Fisch & Spehlmann’s EEG Primer,” -3rd ed. Elsevier, 1999.
[4]Richard Caton. “The electric currents of the brain,” British Medical Journal., 2: 278, 1875.
[5]Berger, H. “ber das elektrenkephalogramm des menschen,” Arch. f. Psychiat., 87:527-570, 1929.
[6]M. Middendorf, G. McMillan, G. Calhoun, and K. S. Jones, “Brain-computer interfaces based on the steady-state visual-evoked response,” IEEE Trans Rehabil Eng, vol. 8, 211-214, 2000.
[7]G. Pfurtscheller, et al. “Graz-BCI: State of the Art and Clinical Applications,” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 11, no. 2, 1 – 4, 2003.
[8]Mark Polak, IEEE Student Member, and Aleksandar Kostov, IEEE Member. “Feature extraction in development of brain-computer interface: a case study,” IEEE in Medicine and Biology Society, vol. 20, no4. 1998.
[9]Jaime A. Pineda, David S. Silverman, Andrey Vankov, and John Hestenes. “Learning to Control Brain Rhythms: Making a Brain-Computer Interface Possible.” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 11, no. 2, 181 – 184, 2003.
[10]E. Donchin, K. M. Spencer, and R. Wijesinghe. “The mental prosthesis: assessing the speed of a P300-based brain- computer interface.” IEEE Trans Rehabil Eng, vol. 8, 174-179, 2000.

[11]G. Pfurtscheller. “Graphical display and statistical evaluation of event-related desynchronization (ERD).” Electroenceph. Clin. Neurophysiol., vol. 43, 757-760. 1977.
[12]G. Pfurtscheller and A. Berghold. “Patterns of cortical activation during planning of voluntary movement.” Electroenceph. Clin. Neurophysiol., vol. 72, 250–258. 1989.
[13]L. Lehtela, R. Salmelin and R. Hari. “Evidence for reactive magnetic 10-Hz rhythm in the human auditory cortex.” Neuroscience Letters, vol. 222, no.2, 111-114. 1997.
[14]Chi Way Chen; Ming Shuang Ju; Lin C.-C.K. “Real time identification of /spl mu/ wave with wavelet neural networks” Neural Engineering, 2003. Conference Proceedings. Page(s):218 – 220, 2003
[15]Klaus-Robert Mller, Charles W. Anderson, and Gary E. Birch. “Comparison of Linear, Nonlinear, and Feature Selection Methods for EEG Signal Classification.” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 11, no. 2, 141 – 144. 2003.
[16]Matthias Kaper, et al. “BCI Competition 2003—Data Set IIb: Support Vector Machines for the P300 Speller Paradigm.” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 51, no. 6, 1073 – 1076. 2004.
[17]Ruei-Chang Wu. “Estimating Driving Performance Based on EEG Spectrum and Fuzzy Neural network.” 2004 IEEE International Joint Conference on Volume 1, 25-29. 2004.
[18]J.-S.R.Jang, C.-T.Sun, E.Mizutani. “Neuro-Fuzzy and Soft Computing.” Prentice-Hall International, Inc. 1997.
[19]http://faculty.washington.edu/chudler/1020.html
[20]Serby, H.; Yom-Tov, E.; Inbar, G.F. ”An Improved P300-Based Brain-Computer Interface.” Neural Systems and Rehabilitation Engineering, IEEE Transactions on [see also IEEE Trans. on Rehabilitation Engineering] Volume 13, Issue 1, 89 – 98. 2005.