本研究旨在針對具導程I心電圖之智慧錶開發一套單導程心肌梗塞辨識演算法暨模型，透過導程I心電圖訊號所產生之特徵訊號辨識心肌梗塞發生於心臟之實際位置；除了準確判斷心肌梗塞產生之位置外，本研究希望開發之演算法暨模型能在低運算資源的硬體環境中執行，以利於心肌梗塞徵兆產生時能及時提供配戴者判斷是否提早就醫之參考。本研究使用Physikalisch-Technische Bundesanstalt (PTB)開放資料集做為單導程心肌梗塞辨識演算法暨模型開發之測試及驗證資料，該資料集包含健康志願者和11種不同部位心肌梗塞的心電圖訊號，資料集之訊號為完整12導程，取樣頻率為1,000赫茲，本研究僅取收錄之Lead I心電訊號進行演算法暨模型開發。首先我們對導程I心電圖進行訊號前處理，包括雜訊濾除、Z分數正規化，接著透過Pan-Tompkins演算法偵測心電圖訊號中的R波並找出其峰值的時間點作為基準點，再以R波基準點進行訊號分割找出完整ECG波形，其完整波形包含P波、Q波、R波、S波及T波。本研究除了將單一完整波形之R波基準點分割為左右兩邊各別萃取時域與頻域之特徵外，並參考專科醫師臨床經驗進行特徵萃取，特徵包含：P波、Q波、S波、T波及其最低點之振幅，以及自定義之R波區段、Q波區段與S波區段3種不同範圍進行時域與頻域之特徵萃取，共計65個特徵使用於辨識阻塞位置。最後，我們將所有萃取之特徵做Min-Max正規化，以排除彼此振幅差異，並提高模型表現。本研究以四種模型並嘗試多種架構進行辨識性能比較，分別為類神經網路(Backpropagation Neural Network, BPNN)、極限梯度提升(eXtreme Gradient Boosting, XGBoost)、支援向量機(Support Vector Machine, SVM) 和隨機森林(Random Forest, RF)，其中隨機森林與該辨識器參數是透過一自動化機器學習平台(Auto Machine Learning Platform)進行多次嘗試後所提供之最佳辨識器。實驗結果以RF表現為最佳，在辨識11種阻塞位置辨識結果中其準確度、靈敏度、特異性、F1 分數與辨識率依序為99.90 %、99.46 %、99.94 %、99.55 %與99.40 %。若考量演算法運算時間與辨識能力，以XGBoost模型表現為最好，其準確度、靈敏度、特異性、F1 分數與辨識率依序為99.89 %、98.81 %、99.94 %、99.16 %與99.35 %。因為本研究開發之智慧錶的取樣頻率無法達到1,000赫茲，因此透過插值法將心電圖訊號取樣轉變為128赫茲以符合智慧錶取樣頻率，並以綜合表現最好的辨識器-XGBoost模型進行測試，其降頻後結果在10等分交叉驗證下的準確度、靈敏度、特異性、F1 分數與辨識率依序為99.87 %、98.60 %、99.93 %、99.04 %和99.24 %。實驗結果證實：在取樣頻率下降後，模型仍維持良好的表現。本研究提出以下三點結論：首先，透過分割每個完整波形並各別萃取特徵的方式將會使辨識準確率提升。第二點，本研究參考臨床經驗進行特徵萃取，將所萃取之特徵納入辨識模型之輸入後發現能有效提升表現，表示臨床多年累積之判斷經驗可取代深度學習網路要進行大量且耗費運算資源之特徵萃取的計算。最後，比對近年發表文獻所提出之辨識結果，在機器學習相關模型裡，本研究所提出之演算法暨模型有更佳的辨識能力。

The aim of this study is to develop a single-lead myocardial infarction (MI) classification algorithm and model for the smart watch with lead I ECG to identify the actual MI location of the heart through the features generated by the lead I ECG signal. In addition to accurately classifying the MI location, it is hoped that the algorithm and model developed in this study can be executed in the watch hardware environment with low computing resources, so as to provide timely reference for the watch user to seek medical treatment in advance when signs of MI occur. In this study, the Physikalisch-Technische Bundesanstalt (PTB) open dataset was used as the testing and validation data for single-lead MI classification algorithm and model development. The signal in the data set is a full 12-lead with a sampling rate of 1,000 Hz. In this study, only the lead I ECG signals from the dataset were used for algorithm and model development. First, we perform signal preprocessing on the lead I ECG, including noise filtering and Z-score normalization. The R peak in the ECG signal was detected by the Pan-Tompkins algorithm, and then we used the R peak to segment each complete ECG waveform that includes P wave, Q wave, R wave, S wave and T wave. In this study, in addition to dividing the R wave reference point of a single complete waveform into the left and right sides to extract the features in the time domain and frequency domain, and we referred to the clinical experience of specialists to extract the features. The features include: P wave, Q wave, S wave, and the amplitudes from their nadirs, as well as the time-domain and frequency-domain features from three self-defined R section, Q section, and S section. A total of 65 features are used to identify the MI locations. Finally, we performed the min-max normalization of all extracted features to exclude amplitude differences from each other and improve model performance. We compared the classification performance with four models and tried a variety of model architectures, including backpropagation neural networks (BPNN), extreme gradient boosting (XGBoost), support vector machine (SVM) and random forest (RF). Among these models, the RF model with the parameters identified by an auto machine learning platform achieved the best performance. Its accuracy, sensitivity, specificity, F1 score and classification accuracy of 11 kinds of MI locations were 99.90%, 99.46%, 99.94%, 99.55% and 99.40%, respectively. Considering the algorithm computational time, the best performance was achieved by the XGBoost model with its accuracy, sensitivity, specificity, F1 score and classification accuracy being 99.89%, 98.81%, 99.94%, 99.16% and 99.35%, respectively. Because the sampling rate of the smart watch developed in this study cannot reach 1,000 Hz, the ECG signal was down sampled to 128 Hz to match the sampling rate of the smart watch. With the features extracted from the downsampling signal, the accuracy, sensitivity, specificity, F1 score and classification accuracy of the XGBoost model with 10 cross-validation were 99.87%, 98.60%, 99.93%, 99.04% and 99.24%, respectively. The experimental results confirmed that the model still maintained good performance even the sampling rate was reduced. The conclusions of this study are as follows: First, segmenting each complete waveforms and extract features individually can improve the classification accuracy. Second, this study takes clinical experience as a reference for feature extraction that these features into the input of the classifier can effectively improve the performance. This indicates that the clinical experience accumulated over years may be more effective in improving classification accuracy than the features extracted by deep learning networks that require a lot of computing resources. Finally, compared with the classification results proposed in the literature published in recent years, the algorithm and model proposed in this study have better classification accuracy.

[1] “Cardiovascular diseases,” Cardiovascular diseases, Jul. 17, 2022. https://www.who.int/health-topics/cardiovascular-diseases (accessed Jul. 17, 2022).
[2] “「憂!台灣心臟出問題?!」-青壯救『心』7行動,”衛生福利部國民健康署, Dec. 31, 2016. https://www.hpa.gov.tw/Pages/Detail.aspx?nodeid=1135&pid=2978 (accessed Jul. 17, 2022).
[3] “103年健保監理指標監測結果報告書.”衛生福利部中央健康保險署. Accessed: Jul. 17, 2022. [Online]. Available: https://www.google.com/url?client=internal-element-cse&cx=012254495936870409035:lzvyrg0mtim&q=https://www.mohw.gov.tw/dl-36740-0a2a6099-6392-496c-ae99-7eb6d0284c22.html&sa=U&ved=2ahUKEwjF2K_Opv_4AhVnpVYBHcMsAPcQFnoECAEQAQ&usg=AOvVaw31KdNz8BzPotlR9zn0XvrU
[4] “急診就醫人次統計－按性別及年齡別分.”衛生福利部統計處. Accessed: Jul. 17, 2022. [Online]. Available: https://www.mohw.gov.tw/dl-61559-ec2c8650-f6b3-40b3-a804-cf5f2eaf4476.html
[5] R. K. Sidhu, “Association between acute myocardial infarction and periodontitis: a review of the literature,” Journal of the International Academy of Periodontology, vol. 18, no. 1, pp. 23–33, Jan. 2016.
[6] A. Borzecki, D. Zółkowska, and M. Sieklucka-Dziuba, “Life style and the risk of development of circulatory system diseases,” Annales Universitatis Mariae Curie-Sklodowska. Sectio D: Medicina, vol. 57, no. 1, pp. 426–432, 2002.
[7] “急性心肌梗塞年奪逾萬命 3縣市發生率最高 | 心臟血管 | 科別 | 元氣網.” https://health.udn.com/health/story/5977/4730349 (accessed Jul. 05, 2022).
[8] C.H. Lee, C.C. Fang, L.M. Tsai, S.T. Gan, S.H. Lin, and Y.H. Li, “Patterns of acute myocardial infarction in Taiwan from 2009 to 2015,” The American Journal of Cardiology, vol. 122, no. 12, pp. 1996–2004, Dec. 2018.
[9] J. C. McSweeney, M. Cody, and P. B. Crane, “Do you know them when you see them? women’s prodromal and acute symptoms of myocardial infarction,” The Journal of Cardiovascular Nursing, vol. 15, no. 3, pp. 26–38, Apr. 2001.
[10] C. J. Ryan and J. J. Zerwic, “Perceptions of symptoms of myocardial infarction related to health care seeking behaviors in the elderly,” The Journal of Cardiovascular Nursing, vol. 18, no. 3, pp. 184–196, Aug. 2003.
[11] N. Arenja et al., “Prevalence, extent, and independent predictors of silent myocardial infarction,” The American Journal of Medicine, vol. 126, no. 6, pp. 515–522, Jun. 2013.
[12] J. G. Canto et al., “Prevalence, clinical characteristics, and mortality among patients with myocardial infarction presenting without chest pain,” Journal of the American Medical Association, vol. 283, no. 24, pp. 3223–3229, Jun. 2000.
[13] V. Randazzo, J. Ferretti, and E. Pasero, “ECG watch: a real time wireless wearable ECG,” in 2019 IEEE International Symposium on Medical Measurements and Applications (MeMeA), pp. 1–6, Jun. 2019
[14] C. Paech et al., “Accuracy of the Apple watch single-lead ECG recordings in pre-term neonates,” Cardiology in the Young, pp. 1–5, Dec. 2021.
[15] J. Bumgarner, J. Bumgarner, C. Lambert, D. Cantillon, B. Baranowski, K. Wolski, A. Hussein, O. Wazni, B. Lindsay, and K. Tarakji, “Assessing the accuracy of an automated atrial fibrillation detection algorithm using novel smartwatch technology among patients presenting for elective cardioversion,” Journal of the American College of Cardiology, vol. 71, p. A411, Mar. 2018.
[16] E. J. Benjamin et al, “Heart disease and stroke statistics—2018 update: a report from the American Heart Association,” Circulation, vol. 137, no. 12, pp. 67–492, Jun. 2018.
[17] A. Earnest, M. E. Hock Ong, N. Shahidah, W. Min Ng, C. Foo, and D. J. Nott, “Spatial analysis of ambulance response times related to prehospital cardiac arrests in the city-state of Singapore,” Prehospital Emergency Care, vol. 16, no. 2, pp. 256–265, Mar. 2012.
[18] J. P. Pell, J. M. Sirel, A. K. Marsden, I. Ford, and S. M. Cobbe, “Effect of reducing ambulance response times on deaths from out of hospital cardiac arrest: cohort study,” The British Medical Journal, vol. 322, no. 7299, pp. 1385–1388, Jun. 2001.
[19] F. Adnet, R. Renault, P. Jabre, E. Kulstad, M. Galinski, and F. Lapostolle, “Incidence of acute myocardial infarction resulting in sudden death outside the hospital,” Emergency medicine journal, vol. 28, no. 10, pp. 884–886, Oct. 2011.
[20] S. Sans, A. Puigdefábregas, G. Paluzie, D. Monterde, and I. Balaguer-Vintró, “Increasing trends of acute myocardial infarction in Spain: the MONICA-Catalonia study,” European Heart Journal, vol. 26, no. 5, pp. 505–515, Mar. 2005 .
[21] J. S. Skinner, D. O. Staiger, and E. S. Fisher, “Is technological change in medicine always worth it? the case of acute myocardial infarction,” Health Affairs, vol. 25, no. Supplement 1, pp. W34–W47, Jan. 2006.
[22] J. Hogenbirk, L. Liboiron-Grenier, R. Pong, and N. Young, “How can telehomecare support informal care? examining what is known and exploring the potential,” vol. 20, May. 2005. [Online]. Available: https://www.researchgate.net/publication/264045698_How_Can_Telehomecare_Support_Informal_Care_Examining_What_is_Known_and_Exploring_the_Potential.
[23] S. Ram, B. Apduhan, and N. Shiratori, “A machine learning framework for edge computing to improve prediction accuracy in mobile health monitoring,” in Computational Science and Its Applications, Springer International Publishing, 2019, pp. 417–431.
[24] E. J. Topol, “Transforming medicine via digital innovation,” Science Translational Medicine, vol. 2, no. 16, p. 16cm4, Jan. 2010.
[25] I. Azimi, A. Anzanpour, A. M. Rahmani, T. Pahikkala, M. Levorato, P. Liljeberg, and N. Dutt, “HiCH: hierarchical fog-assisted computing architecture for healthcare IoT,” ACM Transactions on Embedded Computing Systems, vol. 16, no. 5s, pp. 1–20, Oct. 2017.
[26] I. Azimi, J. Takalo-Mattila, A. Anzanpour, A. M. Rahmani, J.-P. Soininen, and P. Liljeberg, “Empowering healthcare IoT systems with hierarchical edge-based deep learning,” in 2018 IEEE/ACM International Conference on Connected Health: Applications, Systems and Engineering Technologies (CHASE), Sep. 2018, pp. 63–68.
[27] R. Powers, M. Etezadi-Amoli, E. M. Arnold, S. Kianian, I. Mance, M. Gibiansky, D. Trietsch, A. S. Alvarado, J. D. Kretlow, T. M. Herrington, S. Brillman, N. Huang, P. T. Lin, H. A. Pham, and A. V. Ullal, “Smartwatch inertial sensors continuously monitor real-world motor fluctuations in Parkinson’s disease,” Science Translational Medicine, vol. 13, no. 579, p. eabd7865, Feb. 2021.
[28] R. E. Litan, “Vital signs via broadband: remote health monitoring transmits savings, enhances lives,” OPEN MINDS, Aug. 2008, Accessed: Jul. 05, 2022. [Online]. Available: https://openminds.com/market-intelligence/resources/121508betterhealth/
[29] E. P. Widmaier, H. Raff Strang, Kevin T., and A. J. Vander, “Cardiovascular physiology,” in Vander’s human physiology: the mechanisms of body function, 14th ed. New York: McGraw-Hill Education, 2016.
[30] H. Leutheuser, “Wearable computing applications in ehealth,” University of Erlangen-Nürnberg, Germany, 2019.
[31] U. B. Baloglu, M. Talo, O. Yildirim, R. S. Tan, and U. R. Acharya, “Classification of myocardial infarction with multi-lead ECG signals and deep CNN,” Pattern Recognition Letters, vol. 122, pp. 23–30, May 2019.
[32] “ECG changes in myocardial infarction,” Accessed: Jul. 07, 2022. [Online]. Available: https://www.slideshare.net/AdeboyeOluwajuyitan/ecg-changes-in-myocardial-infarction-85782717
[33] S. Mawri, A. Michaels, J. Gibbs, S. Shah, S. Rao, A. Kugelmass, N. Lingam, M. Arida, G. Jacobsen, I. Rowlandson, K. Iyer, A. Khandelwal, and J. McCord, “The comparison of physician to computer interpreted electrocardiograms on ST-elevation myocardial infarction door-to-balloon times,” Critical Pathways in Cardiology, vol. 15, no. 1, pp. 22–25, Mar. 2016.
[34] S. Banerjee and M. Mitra, “Application of cross wavelet transform for ECG pattern analysis and classification,” IEEE Transactions on Instrumentation and Measurement, vol. 63, no. 2, pp. 326–333, Feb. 2014.
[35] D. Corrado, A. Biffi, C. Basso, A. Pelliccia, and G. Thiene, “12-lead ECG in the athlete: physiological versus pathological abnormalities,” British Journal of Sports Medicine, vol. 43, no. 9, pp. 669–676, Sep. 2009.
[36] L. Vicent and M. Martínez-Sellés, “Electrocardiogeriatrics: ECG in advanced age,” Journal of Electrocardiology, vol. 50, no. 5, pp. 698–700, Oct. 2017.
[37] R. S. Khane, A. D. Surdi, and R. S. Bhatkar, “Changes in ECG pattern with advancing age,” Journal of Basic and Clinical Physiology and Pharmacology, vol. 22, no. 4, pp. 97–101, Sep. 2011.
[38] P. W. Macfarlane, S. C. McLaughlin, B. Devine, and T. F. Yang, “Effects of age, sex, and race on ECG interval measurements,” Journal of Electrocardiology, vol. 27 Suppl, pp. 14–19, 1994.
[39] U. R. Acharya, H. Fujita, S. L. Oh, Y. Hagiwara, J. H. Tan, and M. Adam, “Application of deep convolutional neural network for automated detection of myocardial infarction using ECG signals,” Information Sciences, vol. 415–416, pp. 190–198, Nov. 2017.
[40] A. Darmawahyuni, S. Nurmaini, and Sukemi, “Deep learning with long short-term memory for enhancement myocardial infarction classification,” in 2019 6th International Conference on Instrumentation, Control, and Automation (ICA), Jul. 2019.
[41] W. C. Liu, C. S. Lin, C. S. Tsai, T. P. Tsao, C. C. Cheng, J. T. Liou, W. S. Lin, S. M. Cheng, Y. S. Lou, C. C. Lee, and C. Lin, “A deep learning algorithm for detecting acute myocardial infarction,” EuroIntervention: Journal of EuroPCR in Collaboration with the Working Group on Interventional Cardiology of the European Society of Cardiology, vol. 17, no. 9, pp. 765–773, Oct. 2021.
[42] B. Liu, J. Liu, G. Wang, K. Huang, F. Li, Y. Zheng, Y. Luo, and F. Zhou, “A novel electrocardiogram parameterization algorithm and its application in myocardial infarction detection,” Computers in Biology and Medicine, vol. 61, pp. 178–184, Jun. 2015.
[43] S. Banerjee and M. Mitra, “ECG feature extraction and classification of anteroseptal myocardial infarction and normal subjects using discrete wavelet transform,” in 2010 International Conference on Systems in Medicine and Biology, Dec. 2010, pp. 55–60.
[44] N. A. Bhaskar, “Performance analysis of support vector machine and neural networks in detection of myocardial infarction,” Proceedings of the International Conference on Information and Communication Technologies, ICICT 2014, 3-5 December 2014 at Bolgatty Palace & Island Resort, Kochi, India, vol. 46, pp. 20–30, Jan. 2015.
[45] L. N. Sharma, R. K. Tripathy, and S. Dandapat, “Multiscale energy and eigenspace approach to detection and localization of myocardial infarction,” IEEE Transactions on Biomedical Engineering, vol. 62, no. 7, pp. 1827–1837, Jul. 2015.
[46] A. K. Dohare, V. Kumar, and R. Kumar, “Detection of myocardial infarction in 12 lead ECG using support vector machine,” Applied Soft Computing, vol. 64, pp. 138–147, Mar. 2018.
[47] C. Han and L. Shi, “Automated interpretable detection of myocardial infarction fusing energy entropy and morphological features,” Computer Methods and Programs in Biomedicine, vol. 175, pp. 9–23, Jul. 2019.
[48] U. R. Acharya, H. Fujita, V. K. Sudarshan, S. L. Oh, M. Adam, J. E. W. Koh, J. H. Tan, D. N. Ghista, R. J. Martis, C. K. Chua, C. K. Poo, and R. S. Tang, “Automated detection and localization of myocardial infarction using electrocardiogram: a comparative study of different leads,” Knowledge-Based Systems, vol. 99, pp. 146–156, May 2016.
[49] E. Nemati, M. J. Deen, and T. Mondal, “A wireless wearable ECG sensor for long-term applications,” IEEE Communications Magazine, vol. 50, no. 1, pp. 36–43, Jan. 2012.
[50] F. Sabry, T. Eltaras, W. Labda, K. Alzoubi, and Q. Malluhi, “Machine learning for healthcare wearable devices: the big picture,” Journal of Healthcare Engineering, vol. 2022, p. e4653923, Apr. 2022.
[51] R. LeMoyne, T. Mastroianni, D. Whiting, and N. Tomycz, “Assessment of machine learning classification strategies for the differentiation of deep brain stimulation ‘on’ and ‘off’ status for Parkinson’s disease using a smartphone as a wearable and wireless inertial sensor for quantified feedback,” in Wearable and Wireless Systems for Healthcare II: Movement Disorder Evaluation and Deep Brain Stimulation Systems, R. LeMoyne, T. Mastroianni, D. Whiting, and N. Tomycz, Eds. Singapore: Springer, 2019, pp. 113–126.
[52] O. Gómez-Carmona, D. Casado-Mansilla, F. A. Kraemer, D. López-de-Ipiña, and J. García-Zubia, “Exploring the computational cost of machine learning at the edge for human-centric Internet of things,” Future Generation Computer Systems, vol. 112, pp. 670–683, Nov. 2020.
[53] S. R. Pasadyn, M. Soudan, M. Gillinov, P. Houghtaling, D. Phelan, N. Gillinov, B. Bittel, and M. Y Desai, “Accuracy of commercially available heart rate monitors in athletes: a prospective study,” Cardiovascular Diagnosis and Therapy, vol. 9, no. 4, pp. 379–385, Aug. 2019.
[54] 固緯電子, “AFG-3000系列,” gwinstek.com. https://www.gwinstek.com/zh-TW/products/detail/AFG-3000_Series (accessed Jul. 07, 2022).
[55] P. S. Addison, “Wavelet transforms and the ECG: a review,” Physiological measurement, vol. 26, pp. R155-99, Nov. 2005.
[56] S. Banerjee, R. Gupta, and M. Mitra, “Delineation of ECG characteristic features using multiresolution wavelet analysis method,” Measurement, vol. 45, no. 3, pp. 474–487, Apr. 2012.
[57] J. Pan and W. J. Tompkins, “A real-time QRS detection algorithm,” IEEE Transactions on Biomedical Engineering, vol. BME-32, no. 3, pp. 230–236, Mar. 1985.
[58] R. Ghorbani Afkhami, G. Azarnia, and M. A. Tinati, “Cardiac arrhythmia classification using statistical and mixture modeling features of ECG signals,” Pattern Recognition Letters, vol. 70, pp. 45–51, Jan. 2016.
[59] R. L. Devi and V. Kalaivani, “Machine learning and IoT-based cardiac arrhythmia diagnosis using statistical and dynamic features of ECG,” The Journal of Supercomputing, vol. 76, no. 9, pp. 6533–6544, Sep. 2020.
[60] H. Yamaji, K. Iwasaki, S. Kusachi, T. Murakami, R. Hirami, H. Hamamoto, K. Hina, T. Kita, N. Sakakibara, and T. Tsuji, “Prediction of acute left main coronary artery obstruction by 12-lead electrocardiography. ST segment elevation in lead aVR with less ST segment elevation in lead V(1),” Journal of the American College of Cardiology, vol. 38, no. 5, pp. 1348–1354, Nov. 2001.
[61] T. C. Chan, W. J. Brady, R. A. Harrigan, J. P. Ornato, and P. Rosen, Eds., “Acute myocardial infarction: confounding patterns,” in ECG in Emergency Medicine and Acute Care, Philadelphia: Mosby, 2005, pp. 371–388.
[62] M. L. R. Lu, T. De Venecia, S. Patnaik, and V. M. Figueredo, “Atrial myocardial infarction: a tale of the forgotten chamber,” International Journal of Cardiology, vol. 202, pp. 904–909, Jan. 2016.
[63] T. Tabassum and M. Islam, “An approach of cardiac disease prediction by analyzing ECG signal,” in 2016 3rd International Conference on Electrical Engineering and Information Communication Technology (ICEEICT), Sep. 2016, pp. 1–5.
[64] M. H. Jim, C. W. Siu, O. O. Chan, H. W. Chan, W. L. Lee, and C.P. Lau, “Prognostic implications of PR-segment depression in inferior leads in acute inferior myocardial infarction,” Clinical Cardiology, vol. 29, no. 8, pp. 363–368, 2006.
[65] X. Sun, J. Cai, X. Fan, P. Ha, Y. Xie, J. Chen, Y. Xiao, and Y. James Kang, “Decreases in electrocardiographic R-wave amplitude and QT interval predict myocardial ischemic infarction in rhesus monkeys with left anterior descending artery ligation,” Public Library of Science One, vol. 8, no. 8, p. e71876, Aug. 2013.
[66] K. Bansal, M. Gore, and P. Nalabothu, “Anterior myocardial infarction,” in StatPearls, Treasure Island (FL): StatPearls Publishing, 2022. Accessed: Jul. 05, 2022. [Online]. Available: http://www.ncbi.nlm.nih.gov/books/NBK562234/
[67] B. J. Kenny and K. N. Brown, “ECG T wave,” in StatPearls, Treasure Island (FL): StatPearls Publishing, 2022. Accessed: Jul. 05, 2022. [Online]. Available: http://www.ncbi.nlm.nih.gov/books/NBK538264/
[68] A. H. Kashou, H. Basit, and A. Malik, “ST segment,” in StatPearls, Treasure Island (FL): StatPearls Publishing, 2022. Accessed: Jul. 07, 2022. [Online]. Available: http://www.ncbi.nlm.nih.gov/books/NBK459364/
[69] M. Arif, I. A. Malagore, and F. A. Afsar, “Detection and localization of myocardial infarction using K-nearest neighbor classifier,” Journal of Medical Systems, vol. 36, no. 1, pp. 279–289, Feb. 2012.
[70] M. Kumar, R. B. Pachori, and U. R. Acharya, “Automated diagnosis of myocardial infarction ECG signals using sample entropy in flexible analytic wavelet transform framework,” Entropy, vol. 19, no. 9, Art. no. 9, Sep. 2017.
[71] L. D. Sharma and R. K. Sunkaria, “Myocardial infarction detection and localization using optimal features based lead specific approach,” Innovation and Research in BioMedical engineering, vol. 41, no. 1, pp. 58–70, Feb. 2020.
[72] N. Sinha and A. Das, “Identification and localization of myocardial infarction based on analysis of ECG signal in cross spectral domain using boosted SVM classifier,” IEEE Transactions on Instrumentation and Measurement, vol. 70, pp. 1–9, 2021.
[73] 吳浩愷, “自動化機器學習疾病風險預測模型生成平臺之開發,”國立成功大學電機工程學系碩士班碩士論文, 2021.
[74] K. Vandecasteele, T. D. Cooman, Y. Gu, E. Cleeren, K. Claes, W. V. Paesschen, S. V. Huffel, and B. Hunyadi, “Automated epileptic seizure detection based on wearable ECG and PPG in a hospital environment,” Sensors, vol. 17, no. 10, Art. no. 10, Oct. 2017.
[75] Y. Yoon, J. H. Cho, and G. Yoon, “Non-constrained blood pressure monitoring using ECG and PPG for personal healthcare,” Journal of Medical Systems, vol. 33, no. 4, pp. 261–266, Aug. 2009.