過去15年來，缺血性心臟病一直是世界10大首要死因之一，在2016年世界衛生組織(World Health Organization, WHO)的統計中，一千多萬人口死於該病灶。除了冠狀動脈阻塞造成的風險外，還可能引起許多併發症，包括：心室心律不整、心搏停止、房室傳導阻滯，甚至是心衰竭及心因性猝死；上述併發症危險性雖高，但是可以藉由及早血管重建治療來降低其風險，例如：經皮冠狀動脈介入治療、放置血管支架或非侵入性血栓溶劑。然而，現今用於檢測及定位心肌梗塞的醫療儀器，包括：單光子電腦斷層掃描與心導管檢查等，存在許多潛在缺點，低準確度、昂貴、耗時、侵入性或具有放射性、難以重現性及長期監測等，此外視為黃金標準的檢測方法也不具備遠距居家監測、快篩與早期診斷的能力，有鑑於此市場需求，本研究選用普遍被用於監測心臟電氣活動的12導程心電圖並用於心肌缺血的診斷，在現今的研究中，使用12導程心電圖訊號判斷是否發生心肌缺血已經具有不錯的準確度，然而定位心肌缺血發生的位置仍未被實踐。
臨床上，心肌缺血以及梗塞會造成該區域心肌細胞的跨細胞膜電位值異常，特徵有：延遲激發時間、激發電位值降低、靜止電位值上升、激發動作間期縮短等，而具備這些特徵的跨細胞膜電位的心肌細胞會導致該區域復極化電位梯度值上升，並造成鄰近缺血區域的心電圖導程訊號ST間期電位值上升或下降，以及T波電位值上升甚至是反轉等現象，本研究利用了這種現象的電生理機轉，以電腦模擬建立出擬真的心臟放電模型，並於該模型下模擬心肌缺血可能發生的所有位置與缺血嚴重程度後，使用forward計算得到這些心肌缺血模式下的12導程心電圖訊號並建立龐大的心肌缺血心電圖資料庫，以模擬心臟模型結合具有穩健性的稀疏表示式分類演算法應用於具有高度個體變異性的143位心肌梗塞病患，定位心肌梗塞位置的準確度可以達到0.91，最後將心肌缺血位置與嚴重程度從心臟模型投影到二維圓形平面上，過此技術，本研究視覺化的呈現心肌血流灌注情形並提升12導程心電圖診斷心肌梗塞的能力。

In the past 15 years, ischemic heart disease is one of the world’s biggest killers. According to statistics from the World Health Organization in 2016, 10 million deaths were caused by it. This is not only due to the damage caused by coronary artery occlusion, but also many complications of myocardial ischemia including ventricular arrhythmia, cardiac arrest, atrioventricular block, and even heart failure and sudden cardiac death. Although these complications will cause serious damage, it can be prevented by early revascularization such as percutaneous coronary intervention (PCI), placing vascular stents, or thrombolytic agents. However, several clinical examination methods for diagnosing ischemia, such as single photon emission computed tomography and cardiac catheterization, may have many potential shortcomings, including low accuracy, high expense, time consumption, intrusive, the need of injection of radiocontrast agent, the problem of long-term monitoring, and the difficulty of reproducibility. Moreover, the gold standard examination method lacks of the abilities of remote home monitoring, rapid screening, and early diagnosis. In view of this market demand, the 12-lead electrocardiogram (ECG), which is commonly used to monitor the cardiac electrical activity, was selected and used for the diagnosis of myocardial ischemia. Besides, in the existing researches, the 12-lead ECG signals utilized to identify myocardial ischemia has achieved good accuracy. However, the localization of ischemia based on 12-lead ECG has not been practiced.
In clinic, myocardial ischemia and infarction will cause abnormalities to the waveform of myocardial cells’ transmembrane potentials (TMPs) in that region, including delayed activation time, reduced activation potential, rising resting potential, and reduced activation interval. These abnormal TMPs will increase the gradient of repolarized potentials in the ischemic region, and lead to the elevation or depression of the amplitude of ST segments and T waves, and even the inversion of T waves. Accordingly, based on this phenomenon of the electrophysiological mechanism, this study established a realistic ventricle-thorax anisotropic computer model to simulate the real-world electrical pacing of cardiac. By modifying the waveform of TMPs in different regions with various degrees of modification, we could simulate all possible ischemic regions and severity corresponding to real-world patients. Through forward calculation, we could calculate the 12-lead ECG signals from the ischemic cardiac potentials and established a massive 12-lead ECG ischemia database. The realistic computer cardiac model combined with a robust sparse representation classification algorithm was applied to 143 patients with myocardial infarction, and the accuracy of locating myocardial infarction could reach 0.91. Finally, we projected the severity and ischemic region from ventricular model into a 2D circular plane, which could visually display the perfusion of myocardial cells and improve the ability of 12-lead ECG to diagnose myocardial infarction.

[1] P. G. Steg et al., “Prevalence of anginal symptoms and myocardial ischemia and their effect on clinical outcomes in outpatients with stable coronary artery disease: Data from the international observational clarify registry,” JAMA Intern Med, vol. 174, no. 10, pp. 1651-1659, 2014.
[2] F. M. Fesmire et al., “Usefulness of automated serial 12-lead ecg monitoring during the initial emergency department evaluation of patients with chest pain,” Annals of Emergency Medicine, vol. 31, no. 1, pp. 3-11, Jan. 1998.
[3] W. B. Gibler et al., “A rapid diagnostic and treatment center for patients with chest pain in the emergency department,” Annals of Emergency Medicine, vol. 25, no. 1, pp. 1-8, Jan. 1995.
[4] G. S. Kamath et al., “The utility of 12-lead holter monitoring in patients with permanent atrial fibrillation for the identification of nonresponders after cardiac resynchronization therapy,” Journal of the American College of Cardiology, vol. 53, no. 12, pp. 1050-1055, 2009.
[5] J.-c. Hsieh et al., “A cloud computing based 12-lead ecg telemedicine service,” BM”C medical informatics and decision making, vol. 12, no. 1, pp. 1-12, 2012.
[6] Feng K et al., “Myocardial Infarction Classification Based on Convolutional Neural Network and Recurrent Neural Network,” Applied Sciences, 2019.
[7] H. Wang et al., “Myocardial infarction detection based on multi-lead ensemble neural network,” in 2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2019, pp. 2614-2617: IEEE.
[8] K. Nikus et al., “Updated electrocardiographic classification of acute coronary syndromes,” Current cardiology reviews. vol. 10, no. 3, pp. 229-236, 2014.
[9] R. Firoozabadi et al., “Modeling and classification of the st segment morphology for enhanced detection of acute myocardial infarction,” in 2019 Computing in Cardiology (CinC), 2019, pp. Page 1-Page 4: IEEE.
[10] U. B. Baloglu et al., “Classification of myocardial infarction with multi-lead ecg signals and deep cnn,” Pattern Recognition Letters, vol. 122, pp. 23-30, May. 2019.
[11] W. Liu et al., “Multiple-feature-branch convolutional neural network for myocardial infarction diagnosis using electrocardiogram,” Biomedical Signal Processing and Control, vol. 45, pp. 22-32, 2018.
[12] C. Han et al., “Ml–resnet: A novel network to detect and locate myocardial infarction using 12 leads ecg,” Computer methods and programs in biomedicine, vol. 185, p. 105138, 2020.
[13] Bousseljot, R. et al. “Nutzung der EKG-Signaldatenbank CARDIODAT der PTB über das Internet.” Biomedical Engineering/Biomedizinische Technik 40.s1 (1995): 317-318.
[14] S. Al-Zaiti et al., “Machine learning-based prediction of acute coronary syndrome using only the pre-hospital 12-lead electrocardiogram,” Nature communications, vol. 11, no. 1, pp. 1-10, 2020.
[15] R. L. Lux, “Non‐st‐segment elevation myocardial infarction: A novel and robust approach for early detection of patients at risk,” Journal of the American Heart Association, 2015.
[16] Al‐Zaiti et al., “Clinical utility of ventricular repolarization dispersion for real‐time detection of non‐ST elevation myocardial infarction in emergency departments.” Journal of the American Heart Association 4.7 (2015): e002057.
[17] Z. Zhou et al., “Noninvasive imaging of 3-dimensional myocardial infarction from the inverse solution of equivalent current density in pathological hearts,” IEEE Transactions on Biomedical Engineering, vol. 62, no. 2, pp. 468-476, 2014.
[18] F. Dawoud et al., “Using inverse electrocardiography to image myocardial infarction—reflecting on the 2007 physionet/computers in cardiology challenge,” Journal of electrocardiology, vol. 41, no. 6, pp. 630-635, 2008.
[19] M. Lorange et al., “A computer heart model incorporating anisotropic propagation: I. Model construction and simulation of normal activation,” Journal of electrocardiology, vol. 26, no. 4, pp. 245-261, 1993.
[20] A. Van Oosterom et al., “The effect of torso inhomogeneities on body surface potentials quantified using “tailored” geometry,” Journal of electrocardiology, vol. 22, no. 1, pp. 53-72, 1989.
[21] N. Cedilnik et al., “Fast personalized electrophysiological models from computed tomography images for ventricular tachycardia ablation planning,” EP Europace, vol. 20, no. suppl_3, pp. iii94-iii101, 2018.
[22] T. Yang et al., “Localization of origins of premature ventricular contraction by means of convolutional neural network from 12-lead ecg,” IEEE Transactions on Biomedical Engineering, vol. 65, no. 7, pp. 1662-1671, 2017.
[23] R. M. Shaw et al., “Electrophysiologic effects of acute myocardial ischemia: A theoretical study of altered cell excitability and action potential duration,” Cardiovascular research, vol. 35, no. 2, pp. 256-272, 1997.
[24] Van Oosterom et al., “Genesis of the t wave as based on an equivalent surface source model,” Journal of electrocardiology, vol. 34, pp. 217-228, 2001.
[25] P. G. Steg et al., “Prevalence of anginal symptoms and myocardial ischemia and their effect on clinical outcomes in outpatients with stable coronary artery disease: Data from the international observational clarify registry,” JAMA Intern Med, vol. 174, no. 10, pp. 1651-1659, 2014.
[26] L. Weixue et al., “Microcomputer-based cardiac field simulation model,” Medical and Biological Engineering and Computing, vol. 31, no. 4, pp. 384-387, 1993.
[27] R. Doste et al., “A rule‐based method to model myocardial fiber orientation in cardiac biventricular geometries with outflow tracts,” International journal for numerical methods in biomedical engineering, vol. 35, no. 4, p. e3185, 2019.
[28] B. He et al., “Noninvasive imaging of cardiac transmembrane potentials within three-dimensional myocardium by means of a realistic geometry anisotropic heart model,” IEEE Transactions on Biomedical Engineering, vol. 50, no. 10, pp. 1190-1202, 2003.
[29] J. KUPERSMiTH et al., “Conduction intervals and conduction velocity in the human cardiac conduction system: Studies during open-heart surgery,” Circulation, vol. 47, no. 4, pp. 776-785, 1973.
[30] C. Han et al., “Noninvasive three-dimensional cardiac activation imaging from body surface potential maps: A computational and experimental study on a rabbit model,” IEEE transactions on medical imaging, vol. 27, no. 11, pp. 1622-1630, 2008.
[31] Z. Liu et al., “Noninvasive reconstruction of three-dimensional ventricular activation sequence from the inverse solution of distributed equivalent current density,” IEEE transactions on medical imaging, vol. 25, no. 10, pp. 1307-1318, 2006.
[32] R. C. Barr et al., “Relating epicardial to body surface potential distributions by means of transfer coefficients based on geometry measurements,” IEEE Transactions on biomedical engineering, no. 1, pp. 1-11, 1977.
[33] Y. Yamashita et al., “Source-field relationships for cardiac generators on the heart surface based on their transfer coefficients,” IEEE transactions on biomedical engineering, no. 11, pp. 964-970, 1985.
[34] R. N. Klepfer et al., “The effects of inhomogeneities and anisotropies on electrocardiographic fields: A 3-d finite-element study,” IEEE transactions on bio-medical engineering, vol. 44, no. 8, pp. 706-719, 1997.
[35] G. Fischer et al., “A bidomain model based bem-fem coupling formulation for anisotropic cardiac tissue,” Annals of biomedical engineering, vol. 28, no. 10, pp. 1229-1243, 2000.
[36] D. Wu et al., “An improved method for ecg signal feature point detection based on wavelet transform,” in 2012 7th IEEE Conference on Industrial Electronics and Applications (ICIEA). 2012, pp. 1836-1841: IEEE.
[37] Balda RA et al., “The HP ECG analysis program,” In van Bemmel JH and Willems JL, editors, Trends in Computer-processed Electrocardiograms, pages 197-205. North Holland, Amsterdam, The Netherlands, 1977.
[38] P. W. Macfarlane et al., “Comprehensive electrocardiology,” Springer Science & Business Media, 2010.
[39] J. Wright et al., “Robust face recognition via sparse representation,” IEEE transactions on pattern analysis and machine intelligence, vol. 31, no. 2, pp. 210-227, 2008.
[40] Figueiredo, Mário AT et al., “Gradient projection for sparse reconstruction: Application to compressed sensing and other inverse problems.” IEEE Journal of selected topics in signal processing 1.4 (2007): 586-597.
[41] L. Galeotti et al., “Development of an automated method for display of ischemic myocardium from simulated electrocardiograms,” Journal of electrocardiology, vol. 42, no. 2, pp. 204-212, 2009.
[42] S. H. Jambukia et al., “Classification of ecg signals using machine learning techniques: A survey,” in 2015 International Conference on Advances in Computer Engineering and Applications, 2015, pp. 714-721: IEEE.
[43] J. Sohn et al., “Reconstruction of 12-lead electrocardiogram from a three-lead patch-type device using a lstm network,” Sensors, vol. 20, no. 11, p. 3278, 2020.
[44] Z. Xu et al., “Reconstruction of 12-lead electrocardiogram based on gvm,” in 2018 Sixth International Conference on Advanced Cloud and Big Data (CBD), 2018, pp. 275-280: IEEE.
[45] B. J. Drew et al., “Comparison of a new reduced lead set ecg with the standard ecg for diagnosing cardiac arrhythmias and myocardial ischemia,” Journal of electrocardiology, vol. 35, no. 4, pp. 13-21, 2002.
[46] M. Fereniec et al., “The 64 channel system for high resolution ecg mapping,” in Computers in Cardiology 2001, vol. 28 (Cat. No. 01CH37287), 2001, pp. 513-515: IEEE.
[47] B. Khaddoumi et al., “Body surface ecg signal shape dispersion,” IEEE transactions on biomedical engineering, vol. 53, no. 12, pp. 2491-2500, 2006.
[48] Z. Zhou et al., “Noninvasive imaging of high-frequency drivers and reconstruction of global dominant frequency maps in patients with paroxysmal and persistent atrial fibrillation,” IEEE Transactions on Biomedical Engineering, vol. 63, no. 6, pp. 1333-1340, 2016.
[49] C. Han et al., “Imaging cardiac activation sequence during ventricular tachycardia in a canine model of nonischemic heart failure,” American Journal of Physiology-Heart and Circulatory Physiology, vol. 308, no. 2, pp. H108-H114, 2015.
[50] C. Han et al., “Noninvasive imaging of three-dimensional cardiac activation sequence during pacing and ventricular tachycardia,” Heart Rhythm, vol. 8, no. 8, pp. 1266-1272, 2011.
[51] Barold, S. Serge, and ARY L. GOLDBERGER. “A specific ECG triad associated with congestive heart failure,” Pacing and Clinical Electrophysiology 5.4 (1982): 593-599.