臺灣博碩士論文加值系統

為了要開發一高準確率及低複雜度之心電圖訊號特徵萃取系統，利用與心電圖訊號在波型上相似之基底進行特徵萃取的概念是本論文達到高準確偵測的主要策略，而考慮基線飄移雜訊所帶來的影響並且在特徵萃取過程中解決以省略“基線飄移雜訊去除”的步驟是本論文主要降低複雜度的概念；為了達到這些目標，本論文主要提出賈伯小波轉換、不同尺度高斯匹配分析等方法開發出一無須去除基線飄移之心電圖訊號萃取系統；最後，根據主觀及客觀之實驗結果，本論文所提出之無須去除基線飄移演算法不僅有準確偵測結果外，本演算法也可適用於較多種不同之心電圖訊號。因此，本演算法的應用層面不論對於傳統幫助醫療診斷亦或是現今及未來發展中較新穎的即時診斷、健康雲及健康照護等等之應用皆非常適合；另外，本演算法亦可實現在不同平台上，不管是軟體、硬體或嵌入式系統的平台等，都可節省運算時間或消耗能源。

In order to develop a high accuracy and low complexity ECG feature extraction algorithm, using the concept of waveform similarity measurement among the ECG features and corresponding similar bases is the main strategy for accuracy improvement; considering the influences caused by baseline drift noise and solving them during the processing procedures of feature extraction to omit the step of “baseline drift removal” is the major tactic for complexity reduction. Based on these purposes, the primarily detecting methods including Gabor wavelet transform and matching process using half Gaussian model with various scales are proposed to develop an accurate ECG feature extraction systems without baseline drift removal. Finally, subjective and objective experimental results present that proposed baseline drift removal omission algorithm can not only extract the useful features correctly but also be applicable for widely types of ECG signals. Therefore, the proposed algorithm is suitable for not only traditional medical diagnosis but also more novel applications in real-time, health-cloud, health-care systems, etc. In addition, the proposed algorithm could also be implemented on various platforms such as software, hardware, or embedded systems, which may reduce the computing time or power consumption.

摘 要 i
Abstract iii
誌 謝 v
Table of Contents vii
List of Tables ix
List of Figures x
Chapter 1. Introduction 1
1.1 Introduction 1
1.2 Background Information 2
1.3 Motivation 6
1.4 Organization of this Thesis 8
Chapter 2. Surveys of Feature Extraction on ECG Signals in the Literatures 9
2.1 Preprocessing 9
2.2 Feature Extraction 10
2.2.1 Joint Time-Frequency Analysis 11
2.2.2 Derivative-Based Methods 18
2.2.3 Filter-Bank Methods 19
Chapter 3. Proposed Algorithm 23
3.1 Block diagram of Feature Extraction on ECG Signals 23
3.2 Gabor Wavelet Transform 26
3.3 R Peak Detection 30
3.4 Q, S Peaks and QRSon, QRSoff Detection 34
3.5 P, T Peaks Detection 40
3.6 Pon, Poff, Ton, Toff Detection 43
3.7 Amplitudes and Depths Estimation 46
3.8 Baseline Drift Removal Omission 47
Chapter 4. Experimental Results 51
4.1 Experimental database 51
4.2 Subjective results 52
4.3 Objective results 55
Chapter 5. Conclusions and Future Work 72
5.1 Conclusions 72
5.2 Future Work 73
Reference 74

[1]Acharya, U. Rajendra, et al. “Heart rate variability: a review. Medical and Biological Engineering and Computing 44.12 (2006): 1031-1051.
[2]Sayadi, Omid, and Mohammad B. Shamsollahi. “Model-based fiducial points extraction for baseline wandered electrocardiograms. Biomedical Engineering, IEEE Transactions on 55.1 (2008): 347-351.
[3]Luo, Shen, and Paul Johnston. “A review of electrocardiogram filtering. Journal of electrocardiology 43.6 (2010): 486.
[4]Vullings, Rik, Bert de Vries, and Jan WM Bergmans. “An adaptive Kalman filter for ECG signal enhancement. Biomedical Engineering, IEEE Transactions on58.4 (2011): 1094-1103.
[5]Sun, Yan, Kap Luk Chan, and Shankar Muthu Krishnan. “ECG signal conditioning by morphological filtering. Computers in biology and medicine32.6 (2002): 465-480.
[6]Nie, Xiaoning, and Rolf Unbehauen. “Edge preserving filtering by combining nonlinear mean and median filters. Signal Processing, IEEE Transactions on39.11 (1991): 2552-2554.
[7]Hao, Weituo, Yu Chen, and Yi Xin. “ECG baseline wander correction by mean-median filter and discrete wavelet transform. Engineering in Medicine and Biology Society, EMBC, 2011 Annual International Conference of the IEEE. IEEE, 2011.
[8]Gholam-Hosseini, Hamid, Homer Nazeran, and Karen J. Reynolds. “ECG noise cancellation using digital filters. Bioelectromagnetism, 1998. Proceedings of the 2nd International Conference on. IEEE, 1998.
[9]Ahlstrom, Mark L., and Willis J. Tompkins. “Automated high-speed analysis of Holter tapes with microcomputers. Biomedical Engineering, IEEE Transactions on 10 (1983): 651-657.
[10]Fraden, J., and M. R. Neuman. “QRS wave detection. Medical and Biological Engineering and computing 18.2 (1980): 125-132.
[11]Kohler, B-U., Carsten Hennig, and Reinhold Orglmeister. “The principles of software QRS detection. Engineering in Medicine and Biology Magazine, IEEE21.1 (2002): 42-57.
[12]Gabor, Dennis. “Theory of communication. Part 1: The analysis of information. Electrical Engineers-Part III: Radio and Communication Engineering, Journal of the Institution of 93.26 (1946): 429-441.
[13]Qian, Shie, and Dapang Chen. “Joint time-frequency analysis. Signal Processing Magazine, IEEE 16.2 (1999): 52-67.
[14]Gao, Robert X. Wavelets: Theory and applications for manufacturing. Springer Science+ Business Media, 2011.
[15]Lindeberg, Tony. Scale-space theory in computer vision. Kluwer Academic Print on Demand, 1993.
[16]Lee, Gwo Giun, et al. “Gabor feature extraction for electrocardiogram signals. Biomedical Circuits and Systems Conference (BioCAS), 2012 IEEE. IEEE, 2012.
[17]Rusu, Corneliu. “A short way to compute WT from STFT. Proc. EUSIPCO’96. Vol. 1.
[18]Gonzalez, Rafael C., and E. Richard. “Woods, digital image processing. ed: Prentice Hall Press, ISBN 0-201-18075-8 (2002).
[19]McSharry, Patrick E., et al. “A dynamical model for generating synthetic electrocardiogram signals. Biomedical Engineering, IEEE Transactions on50.3 (2003): 289-294.
[20]Laguna, Pablo, et al. “A database for evaluation of algorithms for measurement of QT and other waveform intervals in the ECG. Computers in Cardiology 1997. IEEE, 1997.
[21]http://www.physionet.org/physiobank/database/html/mitdbdir/mitdbdir.htm
[22]Banerjee, Swati, and Madhuchhanda Mitra. “ECG feature extraction and classification of anteroseptal myocardial infarction and normal subjects using discrete wavelet transform. Systems in Medicine and Biology (ICSMB), 2010 International Conference on. IEEE, 2010.
[23]Sahambi, J. S., S. N. Tandon, and R. K. P. Bhatt. “Using wavelet transforms for ECG characterization. An on-line digital signal processing system. Engineering in Medicine and Biology Magazine, IEEE 16.1 (1997): 77-83.
[24]Hamilton, Pat. “Open source ECG analysis. Computers in Cardiology, 2002. IEEE, 2002.
[25]Mazomenos, E. B., et al. “A Time-Domain Morphology and Gradient based algorithm for ECG feature extraction. Industrial Technology (ICIT), 2012 IEEE International Conference on. IEEE, 2012.
[26]Jane, R., et al. “Evaluation of an automatic threshold based detector of waveform limits in Holter ECG with the QT database. Computers in Cardiology 1997. IEEE, 1997.
[27]董玉京, “怎樣看心電圖, 南山堂出版社。
[28]http://www.nottingham.ac.uk/nursing/practice/resources/cardiology/function/placement_of_leads.php
[29]Chen, Po-Ching, Steven Lee, and Cheng-Deng Kuo. “Delineation of T-wave in ECG by wavelet transform using multiscale differential operator. Biomedical Engineering, IEEE Transactions on 53.7 (2006): 1429-1433.
[30]Bednar, J., and T. Watt. “Alpha-trimmed means and their relationship to median filters. Acoustics, Speech and Signal Processing, IEEE Transactions on 32.1 (1984): 145-153.
[31]http://physionet.org/pn3/mghdb/.