Treffer: [Evaluation of an interpretable 12-lead ECG automatic diagnosis model based on deep feature fusion].
Title:
[Evaluation of an interpretable 12-lead ECG automatic diagnosis model based on deep feature fusion].
Authors:
Lu X; School of Biomedical Engineering, Southern Medical University, Guangzhou 510515, China., Chen H; Guangzhou Nanfang Medical Equipment Comprehensive Testing Co., Ltd., Guangzhou 510515, China., Wu Q; School of Biomedical Engineering, Southern Medical University, Guangzhou 510515, China., Wen Y; School of Biomedical Engineering, Southern Medical University, Guangzhou 510515, China., Liu G; Guangdong Medical Devices Quality Surveillance and Test Institute, Guangzhou 510663, China., Chen C; School of Biomedical Engineering, Southern Medical University, Guangzhou 510515, China.
Source:
Nan fang yi ke da xue xue bao = Journal of Southern Medical University [Nan Fang Yi Ke Da Xue Xue Bao] 2026 Jan 20; Vol. 46 (1), pp. 208-218.
Publication Type:
English Abstract; Journal Article
Language:
Chinese
Journal Info:
Publisher: Nanfang yi ke da xue xue bao bian ji bu Country of Publication: China NLM ID: 101266132 Publication Model: Print Cited Medium: Print ISSN: 1673-4254 (Print) Linking ISSN: 16734254 NLM ISO Abbreviation: Nan Fang Yi Ke Da Xue Xue Bao Subsets: MEDLINE
Imprint Name(s):
Original Publication: Guangzhou : Nanfang yi ke da xue xue bao bian ji bu, 2005-
MeSH Terms:
References:
Berkaya SK, Uysal AK, Gunal ES, et al. A survey on ECG analysis[J]. Biomed Sig Proc Control, 2018, 43:216-35.
Chou's Electrocardiography in Clinical Practice[M]. Elsevier Science Health Science div, 2008.
Hong SD, Zhou YX, Shang JY, et al. Opportunities and challenges of deep learning methods for electrocardiogram data:a systematic review[J]. Comput Biol Med, 2020, 122:103801.
Priori SG, Blomström-Lundqvist C, Mazzanti A, et al. 2015 ESC Guidelines for the management of patients with ventricular arrhythmias and the prevention of sudden cardiac death:The Task Force for the Management of Patients with Ventricular Arrhythmias and the Prevention of Sudden Cardiac Death of the European Society of Cardiology(ESC). Endorsed by:Association for European Paediatric and Congenital Cardiology(AEPC)[J]. Eur Heart J, 2015, 36(41):2793-867.
Hannun AY, Rajpurkar P, Haghpanahi M, et al. Cardiologist-level arrhythmia detection and classification in ambulatory electro-cardiograms using a deep neural network[J]. Nat Med, 2019, 25(1):65-9.
Yang XZ, Zhang XY, Yang MY, et al. 12-Lead ECG arrhythmia classification using cascaded convolutional neural network and expert feature[J]. J Electrocardiol, 2021, 67:56-62.
Vaswani A, Shazeer NM, Parmar N, et al. Attention is all you need[C]// Neural Information Processing Systems., 2017.
Dosovitskiy A, Beyer L, Kolesnikov A, et al. An image is worth 16x16 words:Transformers for image recognition at scale[J]. arxiv preprint arxiv:2010. 11929, 2020.
Zhou HY, Zhang SH, Peng JQ, et al. Informer:beyond efficient transformer for long sequence time-series forecasting[J]. Proc AAAI Conf Artif Intell, 2021, 35(12):11106-15.
Lee J, Yoon W, Kim S, et al. BioBERT:a pre-trained biomedical language representation model for biomedical text mining[J]. Bioinformatics, 2020, 36(4):1234-40.
Zhang SY, Lian C, Xu BR, et al. A token selection-based multi-scale dual-branch CNN-transformer network for 12-lead ECG signal classification[J]. Knowl Based Syst, 2023, 280:111006.
Zheng JW, Zhang JM, Danioko S, et al. A 12-lead electrocardiogram database for arrhythmia research covering more than 10, 000 patients[J]. Sci Data, 2020, 7(1):48.
Petch J, Di S, Nelson W. Opening the black box:the promise and limitations of explainable machine learning in cardiology[J]. Can J Cardiol, 2022, 38(2):204-13.
Zhang DD, Yang S, Yuan XH, et al. Interpretable deep learning for automatic diagnosis of 12-lead electrocardiogram[J]. iScience, 2021, 24(4):102373.
Lundberg SM, Lee SI. A unified approach to interpreting model predictions[C]// Neural Information Processing Systems., 2017.
Reddy L, Talwar V, Alle S, et al. IMLE-net:an interpretable multi-level multi-channel model for ECG classification[C]// 2021 IEEE International Conference on Systems, Man, and Cybernetics(SMC). October 17-20, 2021. Melbourne, Australia. IEEE, 2021:1068-1074.
Liu FF, Liu CY, Zhao LN, et al. An open access database for evaluating the algorithms of electrocardiogram rhythm and morphology abnormality detection[J]. J Med Imaging Hlth Inform, 2018, 8(7):1368-73.
Zhang HZ, Zhao W, Liu S. SE-ECGNet:a multi-scale deep residual network with squeeze-and-excitation module for ECG signal classification[C]// 2020 IEEE International Conference on Bioinformatics and Biomedicine(BIBM). December 16-19, 2020. Seoul, Korea. IEEE, 2020:2685-2691.
He KM, Zhang XY, Ren SQ, et al. Deep residual learning for image recognition[C]// 2016 IEEE Conference on Computer Vision and Pattern Recognition(CVPR). June 27-30, 2016, Las Vegas, NV, USA. IEEE, 2016:770-8.
Poli M, Massaroli S, Nguyen E, et al. Hyena hierarchy:Towards larger convolutional language models[C]// International Conference on Machine Learning. PMLR, 2023:28043-78.
Yang W, Deo R, Guo WS. Functional feature extraction and validation from twelve-lead electrocardiograms to identify atrial fibrillation[J]. Commun Med(Lond), 2025, 5(1):32.
Hochreiter S, Schmidhuber J. Long short-term memory[J]. Neural Comput, 1997, 9(8):1735-80.
Kingma D P, Ba J. Adam:A method for stochastic optimization[J]. arxiv preprint arxiv:1412.6980, 2014.
Barredo Arrieta A, Díaz-Rodríguez N, Del Ser J, et al. Explainable artificial intelligence(XAI):concepts, taxonomies, opportunities and challenges toward responsible AI[J]. Inf Fusion, 2020, 58:82-115.
Sundararajan M, Taly A, Yan Q. Gradients of counterfactuals[J]. arxiv preprint arxiv:1611.02639, 2016.
Sokolova M, Lapalme G. A systematic analysis of performance measures for classification tasks[J]. Inf Process Manag, 2009, 45(4):427-37.
Hwang S, Cha J, Heo J, et al. Multi-label abnormality classification from 12-lead ECG using a 2D residual U-net[C]// ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing(ICASSP). April 14-19, 2024, Seoul, Korea, Republic of. IEEE, 2024:2265-9.
Strodthoff N, Wagner P, Schaeffter T, et al. Deep learning for ECG analysis:benchmarks and insights from PTB-XL[J]. IEEE J Biomed Health Inform, 2021, 25(5):1519-28.
Conen D, Adam M, Roche F, et al. Premature atrial contractions in the general population:frequency and risk factors[J]. Circulation, 2012, 126(19):2302-8.
Joglar JA, Chung MK, Armbruster AL, et al. 2023 ACC/AHA/ACCP/HRS guideline for the diagnosis and management of atrial fibrillation:a report of the American college of cardiology/American heart association joint committee on clinical practice guidelines[J]. Circulation, 2024, 149(1):e1-156.
Ikeda T. Right bundle branch block:current considerations[J]. Curr Cardiol Rev, 2021, 17(1):24-30.
Chou's Electrocardiography in Clinical Practice[M]. Elsevier Science Health Science div, 2008.
Hong SD, Zhou YX, Shang JY, et al. Opportunities and challenges of deep learning methods for electrocardiogram data:a systematic review[J]. Comput Biol Med, 2020, 122:103801.
Priori SG, Blomström-Lundqvist C, Mazzanti A, et al. 2015 ESC Guidelines for the management of patients with ventricular arrhythmias and the prevention of sudden cardiac death:The Task Force for the Management of Patients with Ventricular Arrhythmias and the Prevention of Sudden Cardiac Death of the European Society of Cardiology(ESC). Endorsed by:Association for European Paediatric and Congenital Cardiology(AEPC)[J]. Eur Heart J, 2015, 36(41):2793-867.
Hannun AY, Rajpurkar P, Haghpanahi M, et al. Cardiologist-level arrhythmia detection and classification in ambulatory electro-cardiograms using a deep neural network[J]. Nat Med, 2019, 25(1):65-9.
Yang XZ, Zhang XY, Yang MY, et al. 12-Lead ECG arrhythmia classification using cascaded convolutional neural network and expert feature[J]. J Electrocardiol, 2021, 67:56-62.
Vaswani A, Shazeer NM, Parmar N, et al. Attention is all you need[C]// Neural Information Processing Systems., 2017.
Dosovitskiy A, Beyer L, Kolesnikov A, et al. An image is worth 16x16 words:Transformers for image recognition at scale[J]. arxiv preprint arxiv:2010. 11929, 2020.
Zhou HY, Zhang SH, Peng JQ, et al. Informer:beyond efficient transformer for long sequence time-series forecasting[J]. Proc AAAI Conf Artif Intell, 2021, 35(12):11106-15.
Lee J, Yoon W, Kim S, et al. BioBERT:a pre-trained biomedical language representation model for biomedical text mining[J]. Bioinformatics, 2020, 36(4):1234-40.
Zhang SY, Lian C, Xu BR, et al. A token selection-based multi-scale dual-branch CNN-transformer network for 12-lead ECG signal classification[J]. Knowl Based Syst, 2023, 280:111006.
Zheng JW, Zhang JM, Danioko S, et al. A 12-lead electrocardiogram database for arrhythmia research covering more than 10, 000 patients[J]. Sci Data, 2020, 7(1):48.
Petch J, Di S, Nelson W. Opening the black box:the promise and limitations of explainable machine learning in cardiology[J]. Can J Cardiol, 2022, 38(2):204-13.
Zhang DD, Yang S, Yuan XH, et al. Interpretable deep learning for automatic diagnosis of 12-lead electrocardiogram[J]. iScience, 2021, 24(4):102373.
Lundberg SM, Lee SI. A unified approach to interpreting model predictions[C]// Neural Information Processing Systems., 2017.
Reddy L, Talwar V, Alle S, et al. IMLE-net:an interpretable multi-level multi-channel model for ECG classification[C]// 2021 IEEE International Conference on Systems, Man, and Cybernetics(SMC). October 17-20, 2021. Melbourne, Australia. IEEE, 2021:1068-1074.
Liu FF, Liu CY, Zhao LN, et al. An open access database for evaluating the algorithms of electrocardiogram rhythm and morphology abnormality detection[J]. J Med Imaging Hlth Inform, 2018, 8(7):1368-73.
Zhang HZ, Zhao W, Liu S. SE-ECGNet:a multi-scale deep residual network with squeeze-and-excitation module for ECG signal classification[C]// 2020 IEEE International Conference on Bioinformatics and Biomedicine(BIBM). December 16-19, 2020. Seoul, Korea. IEEE, 2020:2685-2691.
He KM, Zhang XY, Ren SQ, et al. Deep residual learning for image recognition[C]// 2016 IEEE Conference on Computer Vision and Pattern Recognition(CVPR). June 27-30, 2016, Las Vegas, NV, USA. IEEE, 2016:770-8.
Poli M, Massaroli S, Nguyen E, et al. Hyena hierarchy:Towards larger convolutional language models[C]// International Conference on Machine Learning. PMLR, 2023:28043-78.
Yang W, Deo R, Guo WS. Functional feature extraction and validation from twelve-lead electrocardiograms to identify atrial fibrillation[J]. Commun Med(Lond), 2025, 5(1):32.
Hochreiter S, Schmidhuber J. Long short-term memory[J]. Neural Comput, 1997, 9(8):1735-80.
Kingma D P, Ba J. Adam:A method for stochastic optimization[J]. arxiv preprint arxiv:1412.6980, 2014.
Barredo Arrieta A, Díaz-Rodríguez N, Del Ser J, et al. Explainable artificial intelligence(XAI):concepts, taxonomies, opportunities and challenges toward responsible AI[J]. Inf Fusion, 2020, 58:82-115.
Sundararajan M, Taly A, Yan Q. Gradients of counterfactuals[J]. arxiv preprint arxiv:1611.02639, 2016.
Sokolova M, Lapalme G. A systematic analysis of performance measures for classification tasks[J]. Inf Process Manag, 2009, 45(4):427-37.
Hwang S, Cha J, Heo J, et al. Multi-label abnormality classification from 12-lead ECG using a 2D residual U-net[C]// ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing(ICASSP). April 14-19, 2024, Seoul, Korea, Republic of. IEEE, 2024:2265-9.
Strodthoff N, Wagner P, Schaeffter T, et al. Deep learning for ECG analysis:benchmarks and insights from PTB-XL[J]. IEEE J Biomed Health Inform, 2021, 25(5):1519-28.
Conen D, Adam M, Roche F, et al. Premature atrial contractions in the general population:frequency and risk factors[J]. Circulation, 2012, 126(19):2302-8.
Joglar JA, Chung MK, Armbruster AL, et al. 2023 ACC/AHA/ACCP/HRS guideline for the diagnosis and management of atrial fibrillation:a report of the American college of cardiology/American heart association joint committee on clinical practice guidelines[J]. Circulation, 2024, 149(1):e1-156.
Ikeda T. Right bundle branch block:current considerations[J]. Curr Cardiol Rev, 2021, 17(1):24-30.
Contributed Indexing:
Keywords: Hyena Hierarchy Convolution Operator; deep learning; electrocardiogram automatic diagnosis; interpretability of model
Local Abstract: [Publisher, Chinese] 目的 : 提升12导联心电图（ECG）自动诊断的准确性和可信度。 方法 : 提出了一种基于深度特征融合的12导联ECG自动诊断模型（MRHL-ECGNet）。该模型包含多尺度特征提取前端、ResNet-34、全局特征混合模块及时间序列分析模块，首次将Hyena Hierarchy卷积算子应用于12导联心电图自动诊断任务中，以高效捕捉ECG中的长程依赖关系，并显著降低模型计算复杂度。同时采用基于积分梯度（IG）的可解释性分析技术，实现MRHL-ECGNet决策依据可视化。使用CPSC2018数据集对MRHL-ECGNet进行训练和测试，并采用多项定量评价指标与评估实验对MRHL-ECGNet进行全面评估。 结果 : 在测试集上对9种类别ECG的分类任务中，MRHL-ECGNet的准确率、AUC值、F1分数、精确率和召回率分别达到0.972、0.983、0.864、0.873和0.857，均优于其他对比模型，且在GPU上对单样本输出诊断结果所需的时间为0.007s，在CPU上也仅需0.156s，内存占用为67.196MB。 结论 : 本研究提出的MRHL-ECGNet不仅具有卓越的分类性能，还具备轻量化及可解释性的特点，在临床ECG辅助诊断中具有较高的应用价值。.
Local Abstract: [Publisher, Chinese] 目的 : 提升12导联心电图（ECG）自动诊断的准确性和可信度。 方法 : 提出了一种基于深度特征融合的12导联ECG自动诊断模型（MRHL-ECGNet）。该模型包含多尺度特征提取前端、ResNet-34、全局特征混合模块及时间序列分析模块，首次将Hyena Hierarchy卷积算子应用于12导联心电图自动诊断任务中，以高效捕捉ECG中的长程依赖关系，并显著降低模型计算复杂度。同时采用基于积分梯度（IG）的可解释性分析技术，实现MRHL-ECGNet决策依据可视化。使用CPSC2018数据集对MRHL-ECGNet进行训练和测试，并采用多项定量评价指标与评估实验对MRHL-ECGNet进行全面评估。 结果 : 在测试集上对9种类别ECG的分类任务中，MRHL-ECGNet的准确率、AUC值、F1分数、精确率和召回率分别达到0.972、0.983、0.864、0.873和0.857，均优于其他对比模型，且在GPU上对单样本输出诊断结果所需的时间为0.007s，在CPU上也仅需0.156s，内存占用为67.196MB。 结论 : 本研究提出的MRHL-ECGNet不仅具有卓越的分类性能，还具备轻量化及可解释性的特点，在临床ECG辅助诊断中具有较高的应用价值。.
Entry Date(s):
Date Created: 20260116 Date Completed: 20260116 Latest Revision: 20260118
Update Code:
20260118
PubMed Central ID:
PMC12809015
DOI:
10.12122/j.issn.1673-4254.2026.01.23
PMID:
41540708
Database:
MEDLINE
Weitere Informationen
Objectives: To enhance the accuracy and reliability of 12-lead electrocardiogram (ECG) automatic diagnosis.
Methods: Herein we propose a 12-lead ECG automatic diagnosis model based on deep feature fusion (MRHL-ECGNet), which consists of a multi-scale feature extraction front-end, ResNet-34, a global feature mixing module, and a time-series analysis module. The Hyena Hierarchy Convolution Operator was applied to the 12-lead ECG automatic diagnosis task for more efficient capture of long-range dependencies while reducing computational complexity. Integrated Gradients (IG)-based interpretability analysis technology was used to achieve visualization of the decision-making basis of MRHL-ECGNet. The CPSC2018 dataset was used to train and test MRHL-ECGNet, and its performance was assessed using multiple quantitative evaluation indicators and evaluation experiments.
Results: In the 9-class ECG classification task on the test set, MRHL-ECGNet achieved an accuracy of 0.972, an AUC of 0.983, an F1 score of 0.864, a precision of 0.873, and a recall of 0.857, all surpassing other comparative models. This model only took 0.007 s to output a diagnosis for a single sample on a GPU and 0.156 s on a CPU, with a memory footprint of 67.196 MB.
Conclusions: The proposed MRHL-ECGNet model demonstrates excellent classification performance in 12-lead ECG automatic diagnosis with a lightweight design and good interpretability, and thus has great potential for clinical application in ECG-aided diagnosis.