心律失常和心肌梗死诊断中心电图智能分析方法研究综述

doi:10.3778/j.issn.1002-8331.2503-0022

摘要/Abstract

摘要： 心电图是诊断心律失常的金标准，且能诊断心肌梗死，其具有无创、实时和便捷等优点，已被广泛应用于临床中。开展心律失常和心肌梗死诊断中心电图智能分析研究具有重要意义。介绍了常用的心律失常和心肌梗死心电数据库；综述了近三年心电图智能分析中的最新技术，包括人工特征提取、卷积神经网络及其变体、图神经网络、自监督学习、联邦学习、主动学习、确定学习和生成式模型；从心电数据规模、分类模式、模型对比和模型复杂度等方面进行对比分析，并重点分析了不同方法的心电数据需求、优缺点、可解释性和应用场景；总结了现有方法在数据质量与类别不均衡、模型泛化性与可解释性的矛盾、隐私保护与协作效率的权衡、计算资源与临床部署的适配性等方面的不足，并给出了可行的解决方案。

关键词: 心电图（ECG）, 心律失常（MI）, 心肌梗死, 智能诊断, 深度学习

Abstract: Electrocardiogram (ECG) remains the gold standard for diagnosing arrhythmia and myocardial infarction (MI), offering advantages such as non-invasiveness, real-time monitoring, and portability, making it widely used in clinical practice. Research on intelligent analysis of ECG for these conditions holds great significance. Firstly, this paper introduces commonly used ECG databases for arrhythmia and MI. Then it reviews recent advances in ECG intelligent analysis over the past three years, including manual feature extraction, convolutional neural networks and their variants, graph neural networks, self-supervised learning, federated learning, active learning, deterministic learning, and generative model. Subsequently, it conducts an in-depth analysis of these methodologies from perspectives including data scale, classification patterns, model comparisons, and model complexity. The study compares the requirements, advantages, disadvantages, interpretability, and application scenarios of different approaches. Finally, it summarizes existing limitations in areas such as data quality and imbalance issues, conflicts between model generalizability and interpretability, trade-offs between privacy protection and collaborative efficiency, mismatches between computational resources and clinical deployment. Feasible solutions are proposed to address these challenges.

Key words: electrocardiogram (ECG), myocardial infarction (MI), arrhythmia, intelligent diagnosis, deep learning

韩闯, 范宝骐, 余梦瑶, 阙文戈. 心律失常和心肌梗死诊断中心电图智能分析方法研究综述[J]. 计算机工程与应用, 2025, 61(23): 59-71.

HAN Chuang, FAN Baoqi, YU Mengyao, QUE Wenge. Review of Intelligent Analysis Methods for Electrocardiograms in Diagnosis of Arrhythmia and Myocardial Infarction[J]. Computer Engineering and Applications, 2025, 61(23): 59-71.

参考文献

[1] GUPTA P, GUPTA M, KUMAR V. An active learning enhanced data programming (ActDP) framework for ECG time series[J]. Machine Learning: Science and Technology, 2024, 5(3): 035016.
[2] LIU F F, LIU C Y, ZHAO L N, et al. An open access database for evaluating the algorithms of electrocardiogram rhythm and morphology abnormality detection[J]. Journal of Medical Imaging and Health Informatics, 2018, 8(7): 1368-1373.
[3] GOLDBERGER A L, AMARAL L A, GLASS L, et al. Physio-Bank, PhysioToolkit, and PhysioNet: components of a new research resource for complex physiologic signals[J]. Circulation, 2000, 101(23): 215-220.
[4] WAGNER P, STRODTHOFF N, BOUSSELJOT R D, et al. PTB?XL, a large publicly available electrocardiography dataset[J]. Scientific Data, 2020, 7(1): 154.
[5] CAI Z P, MA C Y, LI J Q, et al. Hybrid amplitude ordinal partition networks for ECG morphology discrimination: an application to PVC recognition[J]. IEEE Transactions on Instrumentation and Measurement, 2024, 73: 1-13.
[6] LI J, HUANG L, CAI D J, et al. Efficient ECG classification method for arrhythmia using MODWPT and adaptive incremental broad learning[J]. Biomedical Signal Processing and Control, 2025, 104: 107516.
[7] KIRKBAS A, KIZILKAYA A. Automated ECG arrhythmia classification using feature images with common matrix approach-based classifier[J]. Sensors (Basel), 2025, 25(4): 1220.
[8] DARMAWAHYUNI A, NURMAINI S, TUTUKO B, et al. An improved electrocardiogram arrhythmia classification performance with feature optimization[J]. BMC Medical Informatics and Decision Making, 2024, 24(1): 412.
[9] BISWAKARMA A B, RAHUL J, KURMENDRA K. Automated classification of cardiac arrhythmia using short-duration ECG signals and machine learning[J]. Biomedical Physics & Engineering Express, 2025, 11(2): 025020.
[10] LECUN Y, BENGIO Y, HINTON G. Deep learning[J]. Nature, 2015, 521: 436-444.
[11] QIANG Y P, DONG X D, YANG Y. Automatic detection and localisation of myocardial infarction using multi-channel dense attention neural network[J]. Biomedical Signal Processing and Control, 2024, 89: 105766.
[12] HE Z Y, YUAN S Y, ZHAO J H, et al. A robust myocardial infarction localization system based on multi-branch residual shrinkage network and active learning with clustering[J]. Biomedical Signal Processing and Control, 2023, 80: 104238.
[13] HAN C, SUN J J, BIAN Y N, et al. Automated detection and localization of myocardial infarction with interpretability analysis based on deep learning[J]. IEEE Transactions on Instrumentation and Measurement, 2023, 72: 1-12.
[14] 沈尧, 马小虎, 张扬, 等. 基于深度残差记忆网络的心律失常分类方法[J]. 计算机仿真, 2022, 39(12): 342-346.
SHEN Y, MA X H, ZHANG Y, et al. Classification method of arrhythmia based on deep residual memory network[J]. Computer Simulation, 2022, 39(12): 342-346.
[15] HU R, CHEN J, ZHOU L. A Transformer-based deep neural network for arrhythmia detection using continuous ECG signals[J]. Computers in Biology and Medicine, 2022, 144: 105325.
[16] MENG L X, TAN W J, MA J G, et al. Enhancing dynamic ECG heartbeat classification with lightweight transformer model[J]. Artificial Intelligence in Medicine, 2022, 124: 102236.
[17] XIA Y, XIONG Y Q, WANG K Q. A Transformer model blended with CNN and denoising autoencoder for inter-patient ECG arrhythmia classification[J]. Biomedical Signal Processing and Control, 2023, 86: 105271.
[18] JIN Y R, LIU J L, LIU Y Q, et al. A novel interpretable method based on dual-level attentional deep neural network for actual multilabel arrhythmia detection[J]. IEEE Transactions on Instrumentation and Measurement, 2022, 71: 1-11.
[19] YU J, GAO J, WANG N, et al. Spa-tem MI: a spatial-temporal network for detecting and locating myocardial infarction[J]. IEEE Transactions on Instrumentation and Measurement, 2023, 72: 1-12.
[20] HAMMAD M, EL-LATIFA A A, HUSSAIN A, et al. Deep learning models for arrhythmia detection in IoT healthcare applications[J]. Computers and Electrical Engineering, 2022, 100: 108011.
[21] JAHMUNAH V, NG E Y K, TAN R S, et al. Explainable detection of myocardial infarction using deep learning models with Grad?CAM technique on ECG signals[J]. Computers in Biology and Medicine, 2022, 146: 105550.
[22] LIU K G, LIU T, WEN D W, et al. SRTNet: scanning, reading, and thinking network for myocardial infarction detection and localization[J]. Expert Systems with Applications, 2024, 240: 122402.
[23] KIM N, SEO W, KIM J H, et al. WavelNet: a novel convolutional neural network architecture for arrhythmia classification from electrocardiograms[J]. Computer Methods and Programs in Biomedicine, 2023, 231: 107375.
[24] 李国权, 朱双青, 刘梓潼, 等. 基于CvT-13和多模态图像融合的心电分类算法[J]. 生物医学工程学杂志, 2023, 40(4): 736-742.
LI G Q, ZHU S Q, LIU Z T, et al. Electrocardiogram classification algorithm based on CvT-13 and multimodal image fusion[J]. Journal of Biomedical Engineering, 2023, 40(4): 736-742.
[25] JI C Q, WANG L Y, QIN J, et al. MSGformer: a multi-scale grid transformer network for 12-lead ECG arrhythmia detection[J]. Biomedical Signal Processing and Control, 2024, 87: 105499.
[26] LIU Z Y, DENG Z H, LI S M, et al. A multimodal ECG diagnostic framework based on ECG-QRS knowledge graphs and attention mechanisms: enhancing clinical interpretability of ECG diagnostic classification[J]. Biomedical Signal Processing and Control, 2025, 105: 107623.
[27] KAVAK S, CHIU X D, YEN S J, et al. Application of CNN for detection and localization of STEMI using 12-lead ECG images[J]. IEEE Access, 2022, 10: 38923-38930.
[28] PAN W B, AN Y, GUAN Y X, et al. MCA-Net: a multi-task channel attention network for Myocardial infarction detection and location using 12-lead ECGs[J]. Computers in Biology and Medicine, 2022, 150: 106199.
[29] PRABHAKARARAO E, DANDAPAT S. Multi-scale convolutional neural network ensemble for multi-class arrhythmia classification[J]. IEEE Journal of Biomedical and Health Informatics, 2022, 26(8): 3802-3812.
[30] PARK J, LEE K, PARK N, et al. Self-attention LSTM-FCN model for arrhythmia classification and uncertainty assessment[J]. Artificial Intelligence in Medicine, 2023, 142: 102570.
[31] 王建荣, 尉向前, 辛彬彬, 等. 一种改进U-Net网络的心电图分类算法研究 [J]. 重庆理工大学学报(自然科学), 2024, 38(1): 142-149.
WANG J R, WEI X Q, XIN B B, et al. Research on an electrocardiogram classification algorithm for improving U-Net network[J]. Journal of Chongqing University of Technology (Natural Science), 2024, 38(1): 142-149.
[32] HE C, LIU M, XIONG P, et al. Localization of myocardial infarction using a multi-branch weight sharing network based on 2-D vectorcardiogram[J]. Engineering Applications of Artificial Intelligence, 2022, 116: 105428.
[33] HAN K, WANG Y, GUO J, et al. Vision GNN: an image is worth graph of nodes[C]//Advances in Neural Information Processing Systems, 2022: 8291-303.
[34] QIANG Y P, DONG X D, LIU X L, et al. Conv-RGNN: an efficient convolutional residual graph neural network for ECG classification[J]. Computer Methods and Programs in Biomedicine, 2024, 257: 108406.
[35] WANG D M, HU Q H, CAO C G, et al. PM2ECGCN: parallelized spatial-temporal structures of multi-lead ECG with graph convolution network for multi-center cardiac disease diagnosis[J]. Expert Systems with Applications, 2024, 250: 123869.
[36] DUONG L T, DOAN T T H, CHU C Q, et al. Fusion of edge detection and graph neural networks to classifying electrocardiogram signals[J]. Expert Systems with Applications, 2023, 225: 120107.
[37] GE Z Y, CHENG H Q, TONG Z, et al. A knowledge-driven graph convolutional network for abnormal electrocardiogram diagnosis[J]. Knowledge-Based Systems, 2024, 296: 111906.
[38] MA H, XIA L N. Atrial fibrillation detection algorithm based on graph convolution network[J]. IEEE Access, 2023, 11: 67191-67200.
[39] ANDAYESHGAR B, ABDALI-MOHAMMADI F, SEPAHVAND M, et al. Arrhythmia detection by the graph convolution network and a proposed structure for communication between cardiac leads[J]. BMC Medical Research Methodology, 2024, 24(1): 96.
[40] 贺煜航, 刘棪, 陈刚. 基于自适应图卷积网络的心电图多标签分类模型 [J]. 计算机工程, 2022, 48(12): 261-269.
HE Y H, LIU Y, CHEN G. Multi-label classification model of ECG based on adaptive graph convolution network[J]. Computer Engineering, 2022, 48(12): 261-269.
[41] ZHANG H C, LIU W H, CHANG S, et al. ST-ReGE: a novel spatial-temporal residual graph convolutional network for CVD[J]. IEEE Journal of Biomedical and Health Informatics, 2024, 28(1): 216-227.
[42] HE Z Y, CHEN Y F, YUAN S Y, et al. A novel unsupervised domain adaptation framework based on graph convolutional network and multi-level feature alignment for inter-subject ECG classification[J]. Expert Systems with Applications, 2023, 221: 119711.
[43] 喻云虎, 杨湘, 陈艳红. 嵌入导联上下文编码的图卷积神经网络心律失常分类模型[J]. 计算机工程与应用, 2025, 61(3): 212-222.
YU Y H, YANG X, CHEN Y H. Lead context encoding embedded GCN for arrhythmia classification[J]. Computer Engineering and Applications, 2025, 61(3): 212-222.
[44] CHEN Y T, QIU S, WANG Z L, et al. Multiperceptive region of spatial-temporal graph convolutional shrinkage network for arrhythmia recognition[J]. IEEE Transactions on Instrumentation and Measurement, 2024, 73: 1-11.
[45] CHENG Y H, ZHU W L, LI D Y, et al. Multi-label classification of arrhythmia using dynamic graph convolutional network based on encoder-decoder framework[J]. Biomedical Signal Processing and Control, 2024, 95: 106348.
[46] JING L, TIAN Y. Self-supervised visual feature learning with deep neural networks: a survey[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021, 43(11): 4037-4058.
[47] JIN Y R, LI Z Y, TIAN Y Y, et al. A self-supervised framework for computer-aided arrhythmia diagnosis[J]. Applied Soft Computing, 2024, 164: 112024.
[48] SOLTANIEH S, HASHEMI J, ETEMAD A. In-distribution and out-of-distribution self-supervised ECG representation learning for arrhythmia detection[J]. IEEE Journal of Biomedical and Health Informatics, 2024, 28(2): 789-800.
[49] ZHU X Q, SHI M N, YU X X, et al. Self-supervised inter-intra period-aware ECG representation learning for detecting atrial fibrillation[J]. Biomedical Signal Processing and Control, 2025, 100: 106939.
[50] YUN D, YANG H L, KWON S, et al. Automatic segmentation of atrial fibrillation and flutter in single-lead electrocardiograms by self-supervised learning and Transformer architecture[J]. Journal of the American Medical Informatics Association, 2023, 31(1): 79-88.
[51] YANG S X, LIAN C, ZENG Z G, et al. Masked self-supervised ECG representation learning via multiview information bottleneck[J]. Neural Computing and Applications, 2024, 36(14): 7625-7637.
[52] LIU Z Y, LIN C H, HSU Y C, et al. Universal representations in cardiovascular ECG assessment: a self-supervised learning approach[J]. International Journal of Medical Informatics, 2025, 195: 105742.
[53] 陈鹏, 邓淼磊, 樊好义, 等. 心电特征引导下的自监督房颤异常检测方法[J]. 计算机工程与应用, 2025, 61(2): 208-218.
CHEN P, DENG M L, FAN H Y, et al. Self-supervised atrial fibrillation anomaly detection method guided by electrocardiogram features[J]. Computer Engineering and Applications, 2025, 61(2): 208-218.
[54] KONECNY J, MCMAHAN H B, RAMAGE D, et al. Federated optimization: distributed machine learning for on-device intelligence[J]. arXiv:1610.02527, 2016.
[55] RAZA A, TRAN K P, KOEHL L, et al. Designing ECG monitoring healthcare system with federated transfer learning and explainable AI[J]. Knowledge-Based Systems, 2022, 236: 107763.
[56] ASIF R N, DITTA A, ALQUHAYZ H, et al. Detecting electrocardiogram arrhythmia empowered with weighted federated learning[J]. IEEE Access, 2024, 12: 1909-1926.
[57] RAJAGOPAL S M, SUPRIYA M, BUYYA R. FedSDM: federated learning based smart decision making module for ECG data in IoT integrated Edge-Fog-Cloud computing environments[J]. Internet of Things, 2023, 22: 100784.
[58] YING Z B, ZHANG G Y, PAN Z J, et al. FedECG: a federated semi-supervised learning framework for electrocardiogram abnormalities prediction[J]. Journal of King Saud University-Computer and Information Sciences, 2023, 35(6): 101568.
[59] CHORNEY W, WANG H F. Towards federated transfer learning in electrocardiogram signal analysis[J]. Computers in Biology and Medicine, 2024, 170: 107984.
[60] ALRESHIDI F S, ALSAFFAR M, CHENGODEN R, et al. Fed-CL—an atrial fibrillation prediction system using ECG signals employing federated learning mechanism[J]. Scientific Reports, 2024, 14(1): 21038.
[61] WEIMANN K, CONRAD T O F. Federated learning with deep neural networks: a privacy-preserving approach to enhanced ECG classification[J]. IEEE Journal of Biomedical and Health Informatics, 2024, 28(11): 6931-6943.
[62] TANG R J, LUO J Z, QIAN J B, et al. Personalized federated learning for ECG classification based on feature alignment[J]. Security and Communication Networks, 2021, 2021: 6217601.
[63] JOSHI A J, PORIKLI F, PAPANIKOLOPOULOS N. Multi-class active learning for image classification[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2009: 2372-2379.
[64] WANG H Y, ZHOU Y J, NIU X D, et al. A low-cost and stable DAL arrhythmia detection algorithm based on the weak stratification query strategy of morphological statistical features[J]. Expert Systems with Applications, 2025, 261: 125482.
[65] PASOLLI E, MELGANI F. Active learning methods for electrocardiographic signal classification[J]. IEEE Transactions on Information Technology in Biomedicine, 2010, 14(6): 1405-1416.
[66] HE Z Y, YUAN Z Y, AN P F, et al. MFB-LANN: a lightweight and updatable myocardial infarction diagnosis system based on convolutional neural networks and active learning[J]. Computer Methods and Programs in Biomedicine, 2021, 210: 106379.
[67] HE Z Y, YUAN S Y, ZHAO J H, et al. A novel myocardial infarction localization method using multi-branch DenseNet and spatial matching-based active semi-supervised learning[J]. Information Sciences, 2022, 606: 649-668.
[68] FAN W, YANG W Y, CHEN T, et al. AOCBLS: a novel active and online learning system for ECG arrhythmia classification with less labeled samples[J]. Knowledge-Based Systems, 2024, 304: 112553.
[69] WANG C, HILL D J. Learning from neural control[J]. IEEE Transactions on Neural Networks, 2006, 17(1): 130-146.
[70] SUN Q H, LIANG C M, CHEN T R, et al. Early detection of myocardial ischemia in 12‐lead ECG using deterministic learning and ensemble learning[J]. Computer Methods and Programs in Biomedicine, 2022, 226: 107124.
[71] DONG X D, WANG C, SI W J. ECG beat classification via deterministic learning[J]. Neurocomputing, 2017, 240: 1-12.
[72] 孙庆华, 王磊, 王聪, 等. 基于确定学习及心电动力学图的心肌缺血早期检测研究[J]. 自动化学报, 2020, 46(9): 1908-1926.
SUN Q H, WANG L, WANG C, et al. Study on early detection of myocardial ischemia based on definite learning and electrocardiodynamic charts [J]. Automation News, 2020, 46(9): 1908-1926.
[73] SUN Q H, WANG L, LI J L, et al. Multi-phase ECG dynamic features for detecting myocardial ischemia and identifying its etiology using deterministic learning[J]. Biomedical Signal Processing and Control, 2024, 88: 105498.
[74] ZENG W, YUAN C Z. Myocardial infarction detection using ITD, DWT and deterministic learning based on ECG signals[J]. Cognitive Neurodynamics, 2023, 17(4): 941-964.
[75] ZENG W, SHAN L M, YUAN C Z, et al. Detection of myocardial infarction using Shannon energy envelope, FA-MVEMD and deterministic learning[J]. Complex & Intelligent Systems, 2024, 10(4): 4755-4773.
[76] SUN Q H, LI J L, LIANG C M, et al. A multi-lead group network for myocardial infarction detection and localization based on clinical knowledge-driven and dynamic-static feature fusion[J]. Expert Systems with Applications, 2025, 274: 126901.
[77] GOODFELLOW I, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial nets[C]//Proceedings of the 28th International Conference on Neural Information Processing Systems, 2014:2672-2680.
[78] XIA Y, WANG W Y, WANG K Q. ECG signal generation based on conditional generative models[J]. Biomedical Signal Processing and Control, 2023, 82: 104587.
[79] ZHOU F Y, LI J J. ECG data enhancement method using generate adversarial networks based on Bi-LSTM and CBAM[J]. Physiological Measurement, 2024, 45(2): 025003.
[80] 秦静, 韩悦, 王立永, 等. 基于GAN和MS-ResNet的房颤自动检测模型[J]. 应用科学学报, 2024, 42(1): 15-26.
QIN J, HAN Y, WANG L Y, et al. Automatic detection model of atrial fibrillation based on GAN and MS-ResNet[J]. Journal of Applied Science, 2024, 42(1): 15-26.
[81] ALCARAZ J M L, STRODTHOFF N. Diffusion-based conditional ECG generation with structured state space models[J]. Computers in Biology and Medicine, 2023, 163: 107115.
[82] ADIB E, FERNANDEZ A S, AFGHAH F, et al. Synthetic ECG signal generation using probabilistic diffusion models[J]. IEEE Access, 2023, 11: 75818-75828.
[83] ZAMA M H, SCHWENKER F. ECG synthesis via diffusion-based state space augmented transformer[J]. Sensors (Basel), 2023, 23(19): 8328.