利用數位訊號處理與智慧型理論針對正常人初期睡眠之腦波分析研究

The Analysis of Adults’ Brain Wave in Early Sleep Using Digital Signal Processing and Fuzzy Set Theory

莊承澤(Chen-Zer Chuang)

指導教授 ： 謝建興

元智大學/工程學院/機械工程學系/碩士(2001年)

摘要

本論文是利用非週期波理論、95%頻譜邊緣頻率分析、複雜度等數位訊號處理方法做腦波訊號處理與特徵截取。之後，再利用模糊理論建立一「智慧型腦波辨識系統」，分析正常人由清醒進入睡眠初期階段，藉由定量的參數輸入而劃分出意識程度。本研究分別分析F8-T4、P4-O2、C4-F4、C4-T4和C4-P4五點的量測位置，結果發現C4-T4為最不受干擾的量測位置，故訊號處理在C4-T4也為最精確，以此位置建立腦波辨識系統為最佳。未來若欲達到程度60的辨識要加入時變量才可，同時其他的位置都有嚴重的訊號干擾，必須先將其外界干擾去除才可做腦波分析。

關鍵字

非週期波理論 ； 95%頻譜邊緣頻率分析 ； 複雜度 ； 模糊理論

並列摘要

In this study, we used the aperiodic theory, 95% spectrum edge frequency, and complexity analysis to do brain wave signal processing. And we used the fuzzy theory to construct “An Intelligent Identification System of Brain Wave Signals”. The system is used quantitative parameters to analyse the stage from awake to early sleep. After the analyses of the five positions of the F8-T4, P4-O2, C4-F4, C4-T4, and C4-P4, this research was found that C4-T4 has less noise interferences. Also, signal processing in C4-T4 is best than other positions. However, regarding identified conscious level 60 of the C4-T4, it still needs to add a time-change parameter. It is important to know that having a good brain signal processing results must avoid any interference during data acquisition.

並列關鍵字

Aperiodic Theory ； 95% Spectrum Edge Frequency ； Complexity ； Fuzzy Theory

參考文獻

[3] Gubner J.A., Chang W.B., “Wavelet transforms for discrete-time periodic signals”, Signal Processing, Vol. 42, 1995, Page(s): 167-180

[4] Geva A.B., Kerem D.H., “Forecasting Generalized Epileptic seizures from the EEG signal by wavelet analysis and dynamic unsupervised fuzzy clustering”, IEEE Transactions On Biomedical Engineering, Vol. 45, No.10, 1998, Page(s): 1205-1216

[5] Glass PS, Bloom M, Kearse L,” Bispectral analysis measures sedation and memory effects of propofol, midazolam, isoflurane, and alfentanil in healthy volunteers”, Anesthesiology, Vol.86, No.4, 1997, Page(s): 836-847

[6] Sclabassi R.J., Sun M., Qian S. “Localizing Functional Activity in the Brain Through Time-Frequency Analysis and Synthesis of the EEG”, Proceedings of the IEEE, Vol.84, No.9, 1996, Page(s): 1302-1311

[7] Pezard L., Jech R., Ruzicka E., “Investigation of non-linear properties of multichannel EEG in the early stage of Parkinson’s disease”, Clinical Neurophysiology, Vol. 112, 2001, Page(s): 38-45