基于迁移学习的心电图分类识别算法的研究

doi:10.11896/jsjkx.240900073

摘要/Abstract

摘要： 随着城市生活节奏的不断加快,越来越多的人被心血管疾病所困扰。心电图作为诊断心脏病的关键手段,在面对日益增长的患者数量时,有限的医疗资源难以满足庞大的心电图判读需求。因此,如何利用计算机自动分类识别心电图成为了一个迫切的需求。文中基于安贞医院提供的临床数据集,经统计该数据集中存在着数据总量少、数据分布不均匀、部分数据未标注的问题。基于此,使用半监督学习方法对未标注数据进行标注,算法标注精度达到了91.4%。其次,使用迁移学习对模型进行训练,所用源数据集和目标数据集的MMD值为1.99,两者分布有着较高的相似度,与其他训练方法相比,该算法能够在数据总量较小且数据分布不均匀的数据集上取得较好的学习效果;在实际的门诊数据集上,该方法使模型的精确度达到了0.973,召回率达到了0.866,F1值达到了0.932,与不使用迁移学习相比,精确度提升了0.423,召回率提升了0.274,F1值提升了0.384。这一结果表明,该算法具有较好的泛化能力和适应性,可为临床实践提供有力的支持。

关键词: 心电图, 迁移学习, 半监督学习

Abstract: As the pace of urban life continues to accelerate,more and more people are troubled by cardiovascular diseases.Electrocardiogram is a key means of diagnosing heart disease,but in the faced of the growing number of patients,limited medical resources cannot meet the huge demand for electrocardiogram interpretation.Therefore,how to use computers to automatically classify and identify electrocardiograms has become an urgent need.This study is based on the clinical data set provided by Anzhen hospital.According to statistics,there are problems in the data set such as small total data,uneven data distribution,and some data are not labeled.Based on this,this paper uses a semi-supervised learning method to label the unlabeled data,and the algorithm labeling accuracy reaches 91.4%.Secondly,this paper uses transfer learning to train the model.The MMD value of the source data set and the target data set used in this paper is 1.99,and the distribution of the two has a high similarity.Compared with other training methods,this algorithm can achieve better learning results on data sets with a small total amount of data and uneven data distribution; on the actual outpatient data set,our method makes model’s accuracy reach 0.973,recall rate reaches 0.866,and F1 value reaches 0.932.Compared with not using transfer learning,accuracy is improved by 0.423,recall rate is improved by 0.274,and F1 value is improved by 0.384.This results show that the algorithm has good generalization ability and adaptability,and can provide strong support for clinical practice.

Key words: ECG, Transfer learning, Semi-supervised learning

中图分类号:

TP312

陈麒瑞, 王宝会, 戴辰程. 基于迁移学习的心电图分类识别算法的研究[J]. 计算机科学, 2025, 52(6A): 240900073-8. https://doi.org/10.11896/jsjkx.240900073

CHEN Qirui, WANG Baohui, DAI Chencheng. Research on Electrocardiogram Classification and Recognition Algorithm Based on Transfer Learning[J]. Computer Science, 2025, 52(6A): 240900073-8. https://doi.org/10.11896/jsjkx.240900073

参考文献

[1]HANNUNA Y,RAJPURKAR P,HAGHPANAHI M, et al.Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network[J/OL].Nat Med,2019,25:65-69.https://doi.org/10.1038/s41591-018-0268-3.
[2]IVORA A,VISCOR I,NEJEDLY P.et al.QRS detection and classification in Holter ECG data in one inference step[J].Sci Rep,2022,12,12641.
[3]JIANG M M,YU B,ZHANG S,et al.Electrocardiogram Ana-lysis Based on Multiscale Fast Sample Entropy and Random Forest[J].Progress in Modern Biomedicine,2018,8(18):3453-3458.
[4]ZHOU J.Research on ECG abnormality diagnosis based on deep transfer learning[D].Qilu University of Technology.
[5]WANG X,ZHU E Z,AI Z S.Principle of convolutional neural network and its applications in medical imaging diagnosis[J].Chinese Journal of Medical Physics,2022,39(12):1485-1489.
[6]LIMA E M,RIBEIRO A H,PAIXÃO G M M,et al.Deep neural network-estimated electrocardiographic age as a mortality predictor[J].Nat Commun,2021,12(1):5117.
[7]WEIMANN K,CONRAD T O F.Transfer learning for ECGclassification[J].Sci Rep,2021,11,5251.
[8]HE K,ZHANG X,REN S,et al.Deep Residual Learning for Image Recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition(CVPR).Las Vegas,NV,USA,2016:770-778.
[9]SALEEM S,KHANDOKER A H,ALKHODARI M,et al.Atwo-step pre-processing tool to remove Gaussian and ectopic noise for heart rate variability analysis[J].Sci Rep,2022,12,18396.
[10]LI J,PANG S P,XU F.et al.Two-dimensional ECG-based cardiac arrhythmia classification using DSE-ResNet[J].Sci Rep,2022,12,14485.
[11]EONG D U,LIM K M.Combined deep CNN-LSTM network-based multitasking learning architecture for noninvasive continuous blood pressure estimation using difference in ECG-PPG features[J].Sci Rep,2021,11,13539.
[12]RIBEIRO A H,RIBEIRO M H,PAIXÃO G M M,et al.Automatic diagnosis of the 12-lead ECG using a deep neural network[J].Nat Commun,2020,11,1760.
[13]ALARSAN F I,YOUNES M.Analysis and classification ofheart diseases using heartbeat features and machine learning algorithms[J].J Big Data,2019,6:81.
[14]XU J Q,YUAN Z D,LI G P,et al.12-lead electrocardiogram feature extraction method based on wavelet transform [J]. Journal of East China Normal University (Natural Science Edition),2005(2):27-32.
[15]ZHAO C M,MA X H.Electrocardiogram classification based on non-negative matrix decomposition and support vector machine [J].Computer Engineering,2012,38(9):174-176.
[16]JIANG M M,YU B,ZHANG S,et al.Electrocardiogram analysis based on multi-scale fast sample entropy and random forest [J].Progress in Modern Biomedicine,2018,18(18):3453-3458.
[17]CHENG Z Y,ZHANG B L,CAI Y C,et al.Research on prediction and feature analysis of heart disease by integrating random forest and SHAP[J].Intelligent Computers and Applications,2023,13(11):172-179.
[18]LIU Y Q,SONG Y,LI Z M,et al.ECG signal analysis and detection system based on deep learning[J].Journal of Jilin University(Information Science Edition),2023,41(6):1135-1142.
[19]GAO X R,CAO J D,ZHANG J R.An ECG classification method based on deep learning[J].Journal of Nantong University(Natural Science Edition),2023,22(3):17-25.
[20]WEISS K,KHOSHGOFTAAR T M,WANG D D.A survey of transfer learning[J].Journal of Big Data,2016,3(1):1-40.
[21]SU H X,XU X L.Application analysis of deep learning in electrocardiogram classification[J].Biomedical Engineering Research,2020,39(4):419-424.
[22]WANG X,ZHU E Z,AI Z S.Principles of convolutional neural networks and their applications in medical imaging diagnosis [J].Chinese Journal of Medical Physics,2022,39(12):1485-1489.
[23]WANG D H.Multivariate statistical analysis and SPSS application(2nd edition)[M].Shanghai:East China University of Science and Technology Press.2018.02.9G10.
[24]ARTHUR G,KARSTEN M B,MALTE R,et al.A kernelmethod for the two-sample-problem[C]//Proceedings of the Advances in Neural Information Processing Systems.2007:513-520.
[25]余圣波.基于样本差异的多标签迁移学习[D].重庆大学,2016.
[26]CHAPELLE O,SCHOLKOPF B, ZIEN A,et al.Semi-Super-vised Learning[C]//IEEE Transactions on Neural Networks.2009:542-542.
[27]WAGSTAFF K,CARDIE C,ROGERS S,et al.Constrained K-means Clustering with Background Knowledge[C]//Proceedings of the Eighteenth International Conference on Machine Learning.2001:577-584.
[28]FAZAKIS N,KARLOS S,KOTSIANTIS S,et al.A Multi-scheme Semi-supervised Regression Approach[J].Pattern Recognition Letters,2019,125:758-765.
[29]KOSTOPOULOS G,KARLOS S,KOTSIANTIS S,et al.Semi-supervised Regression:A Recent Review[J].Journal of Intelligent & Fuzzy Systems,2018,35(2):1483-1500.
[30]TRIGUERO I,SAEZ J A,LUENGO J,et al.On the Characterization of Noise Filters for Self-training Semi-supervised in Nearest Neighbor Classification[J].Neurocomputing,2014,132:30-41.
[31]SEOK K.Semi-supervised Regression Based on Support Vector Machine[J].Journal of the Korean Data and Information Science Society,2014,25(2):447-454.
[32]OQUAB M,BOTTOU L,LAPTEV I,et al.Learning and transferring mid-level image representations using convolutional neural networks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2014.
[33]LIU H,WANG Y L,CHEN D,et al.A large-scale multi-label 12-lead electrocardiogram database with standardized diagnostic statements.figshare.Collection[J].https://doi.org/10.6084/m9.figshare.c.5779802.v1,2022.
[34]CLIFFORDG D,LIU C Y,MOOD Y B,et al.Af classificationfrom a short single lead ecg recording[J].The Physionet Computing in Cardiology Challenge,2017.
[35]ZHANG D Y,GENG S J,ZHOU Y,et al.Artificial Intelligence System for Detection and Screening of Cardiac Abnormalities using Electrocardiogram Images[J].arXiv:2302.10301,2023.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed