模糊小腦模型類神經網路應用於自動睡眠階段分類與預測

本論文提出一個針對腦電圖的睡眠分類與嗜睡預測使用了希爾伯特 - 黃變換和模糊小腦模型類神經網路。由於腦電圖是一種複雜且非線性的訊號，所以本論文利用希爾伯特 - 黃變換來分析腦電波訊號，因為希爾伯特-黃變換引入了基於信號局部特徵的固有模態函數，因此適用於分析非線性、非平穩訊號，進而求出具有物理意義的瞬時頻率，並得到腦電訊號在頻率上的能量分布，作為睡眠腦電波各個階段的特徵，再利用本質模態函數分析單位訊號中的睡眠紡錘波的多寡，而得到具有判別能力的特徵，採用模糊小腦模型類神經網路來進行分類與預測，使用導傳遞來訓練模糊小腦模型類神經網路的參數，使得此分類器具有自動判別睡眠階段與嗜睡預測的功能。模擬結果顯示，此方法可以有效且快速的利用單一腦電波進行分類預測，且每判斷一組訊號的時間不到一秒鐘，所以可以應用於駕駛人嗜睡預測，也能進行即時睡眠品質監測和睡眠有關的疾病診斷。

關鍵字

模糊小腦模型類神經網路；腦電圖；希爾伯特-黃變換；自動睡眠階段分類；嗜睡預測

並列摘要

In this thesis, a novel method, based on Hilbert-Huang Transform (HHT) and Fuzzy Cerebellar Model Neural Network (FCMNN), is proposed to perform automatic sleep stage classification and prediction for electroencephalographic (EEG). Since EEG is a complex and non-linear signals, so HHT is used to analyze the EEG. Because HHT is based on local signal characteristics intrinsic mode function, it is applied to the analysis of nonlinear and non-stationary signals. The instantaneous frequencies of sleep EEG with physical meaning and energy frequency distribution are used as feature parameters for each stage which are were computed with the HHT. A methodology is develope to detect and characterize sleep spindles, based on the intrinsic mode functions (IMF). The sleep spindles quantity can enhance the discriminative ability of features. Finally, FCMNN is used for the classification and prediction of sleep stage. Back-propagation algorithm is used to learn the parameter of FCMNN. The experiment results show the proposed method is effective and fast for a single EEG signal classification and prediction. This algorithm can determine a set of signals with the computation time less than a second. So this method can be applied to the drowsiness prediction of a driver, and it can be also a powerful tool in on-line sleep quality monitoring and sleep-related diseases diagnosis.

並列關鍵字

FCMNN ； EEG ； Hilbert-Huang Transform ； Automatic sleep stage classification ； Drowsiness prediction

參考文獻

[1] A.L. Loomis, E.N. Harvey, and G.A Hobart, “Cerebral States During Sleep, as Studied by Human Brain Potentials,” J. Exp. Psychol., Vol. 21, No. 2, pp. 127–144, 1937.

[2] Rechtschaffen and A. Kales, “A Manual of Standardized Terminology, Techniques and Scoring System for Sleep Stages of Human Subjects,” US, Government Printing Office, National Institute of Health Publication, Washington DC, 1969.

[3] P. N. Prinz, L. H. Larsen, K. E. Moe, and M. V. Vitiello, “EEG markers of early Alzheimer’s disease in computer selected tonic REM sleep,” Electroencephalogr. Clin. Neurophysiol., Vol. 83, No. 2, pp. 36–43, 1992.

[6] Oropesa E, Cycon HL, Jobert M, “Sleep Stage Classification using Wavelet Transform and Neural Network,” JCSl Technical Report , TR-99-008, 1999.

[7] LI Yong, ZHANG Shengxun, “Apply Wavelet Transform to analysis EEG signal,” 18th Annual1nternational Conference of the lEEE Engineering In Medicine and Biology Society, Amsterdam. pp. 1007-l008, 1996.

國際替代計量

模糊小腦模型類神經網路應用於自動睡眠階段分類與預測

全文下載

主題瀏覽