计算机科学

计算机科学 ›› 2024, Vol. 51 ›› Issue (12): 174-180.doi: 10.11896/jsjkx.230800083

• 数据库&大数据&数据科学 • 上一篇    下一篇

基于梯度幅值方向调整的心电信号多任务分类算法

张雪1, 田岚1, 曾鸣1, 刘俊晖1, 宗绍国2   

  1. 1 山东大学微电子学院 济南 250101
    2 山东省厚德测控技术股份有限公司 济南 250000
  • 收稿日期:2023-08-15 修回日期:2024-03-01 出版日期:2024-12-15 发布日期:2024-12-10
  • 通讯作者: 田岚(tianlan65@sdu.edu.cn)
  • 作者简介:(202212332@mail.sdu.edu.cn)
  • 基金资助:
    山东省自然科学基金(ZR2021MF065,ZR2021ZD40)

Multitask Classification Algorithm of ECG Signals Based on Radient Magnitude Direction Adjustment

ZHANG Xue1, TIAN Lan1, ZENG Ming1, LIU Junhui 1, ZONG Shaoguo2   

  1. 1 School of Microelectronics, Shandong University, Jinan 250101, China
    2 Shandong Houde Measurement and Control Technology Co., LTD, Jinan 250000, China
  • Received:2023-08-15 Revised:2024-03-01 Online:2024-12-15 Published:2024-12-10
  • About author:ZHANG Xue,born in 2000,postgra-duate,is a member of CCF(No.T3347G).Her main research interests include biological signal processing and intelligent healthcare.
    TIAN Lan,born in 1965,professor,Ph.D supervisor.Her main research interests include brain-computer interface,brain-like computing and deep learning,biological signal processing and light regulation of neural activity.
  • Supported by:
    Natural Science Foundation of Shandong Province,China(ZR2021MF065,ZR2021ZD40).

PDF (PC)

154

摘要： 心血管疾病对人类生命健康安全的威胁日益严重,通过心电信号可进行相关疾病的诊断分类。现有的心电分类算法大多采用单任务学习模型,无法综合利用多个任务中的互补特征,而多任务学习模型可同时学习多个相关任务,共享相关任务特征,有助于提高多任务的分类表现。结合深度学习和多任务学习两种方法,提出了一种基于损失优化的心电信号多任务分类算法,将心电信号的多分类任务分解为多个二分类任务,从任务梯度的幅值和方向两方面进行损失优化,避免手动设置任务损失权重以及任务损失相互抵消而产生的负迁移,从而提升心电信号多分类任务的性能。在PTB-XL数据库上将心电信号23类分类任务分解为23个二分类任务来评估所提出的算法。实验结果表明,所提算法的宏观曲线下平均面积(AUC)达到0.950,准确率达到96.50%,基于标签的宏观F1分数达到0.583,基于样本的F1分数达到0.777。与单任务学习算法相比,所提算法在心电信号的多分类方面表现出良好的性能。

关键词: 心电信号分类, 多任务学习, 损失优化

Abstract: Cardiovascular diseases are posing more and more serious threats to human health and safety.ECG signals can be used to diagnose and classify related diseases.Most existing ECG classification algorithms adopt single-task learning model,which can not make comprehensive use of complementary features in multiple tasks.However,multi-task learning model can learn multiple related tasks at the same time,share related task features,and help improve the classification performance of multiple tasks.Combining deep learning and multi-task learning,a multi-task classification algorithm for ECG signals based on loss optimization is proposed.The multi-classification task of ECG signals is decomposed into multiple binary classification tasks,and loss optimization is carried out from the aspects of the amplitude and direction of task gradient,so as to avoid the negative transfer caused by manual setting of task loss weights and the cancellation of task losses.The performance of ECG signal multi-classification task is improved.The model uses PTB-XL database to decompose 23 classification tasks into 23 binary classification tasks to evaluate the proposed algorithm.Experimental results show that the average area under the macro curve(AUC) reaches 0.950,the accuracy reaches 96.50%,the tag-based F1 score reaches 0.583,and the sample-based F1 score reaches 0.777.Compared with the single-task learning algorithm,the proposedalgorithm shows good performance in the multi-classification of ECG signals.

Key words: ECG signal classification, Multi-task learning, Loss optimization

中图分类号: 

  • TP391.7
[1]BENJAMIN E J,VIRANI S S,CALLAWAY C W,et al.Heart Disease and Stroke Statistics－2018 Update:A Report From the American Heart Association[J].Circulation,2018,137(12):E67-E492.
[2]LUZ E J D,SCHWARTZ W R,CAMARA-CHAVEZ G,et al.ECG-based heartbeat classification for arrhythmia detection:A survey[J].Computer Methods and Programs in Biomedicine,2016,127:144-164.
[3]BERKAYA S K,UYSAL A K,GUNAL E S,et al.A survey on ECG analysis[J].Biomedical Signal Processing and Control,2018,43:216-235.
[4]ZHAO R X,XU P J,LIU Y.ECG-based Atrial Fibrillation Detection Based on Deep Convolutional Residual Neural Network[J].Computer Science,2022,49(5):186-193.
[5]HANNUN A Y,RAJPURKAR P,HAGPANAHI M,et al.Car-diologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network[J].Nature Medicine,2019,25(1):65-69.
[6]AHMED A A,ALI W,ABDULLAH T A A,et al.Classifying Cardiac Arrhythmia from ECG Signal Using 1D CNN Deep Learning Model[J].Mathematics,2023,11,562:1-16.
[7]ACHARYA U R,OH S L,HAGIWARA Y,et al.A deep con-volutional neural network model to classify heartbeats[J].Computers in Biology and Medicine,2017,89:389-396.
[8]YILDIRIM O,PLAWIA P,TAN R S,et al.Arrhythmia detection using deep convolutional neural network with long duration ECG signals[J].Computers in Biology and Medicine,2018,102:411-420.
[9]ULLAH H,HEYAT M,AKHTAR F,et al.An Automatic Premature Ventricular Contraction Recognition System Based on Imbalanced Dataset and Pre-Trained Residual Network Using Transfer Learning on ECG Signal[J].Diagnostics,2023,13(1):1-20.
[10]CARUANA R.Multitask learning[J].Machine Learning,1997,28(1):41-75.
[11]ZHANG Y,YANG Q.An overview of multi-task learning[J].National Science Review,2018,5(1):30-43.
[12]SHAHI M,OO E,AHMED B,et al.Adversarial Multi-TaskLearning for Robust End-to-End ECG-based Heartbeat Classification[C]//Proceedings of the 42nd Annual International Conference of the IEEE Engineering in Medicine and Biology Society.Montreal,Canada,2020:341-344.
[13]ZHANG Y,YANG Q.A Survey on Multi-Task Learning[J].IEEE Transactions on Knowledge and Data Engineering,2022,34(12):1-20.
[14]HUANG Z C,LI W,XIA X G,et al.A Novel Nonlocal-Aware Pyramid and Multiscale Multitask Refinement Detector for Object Detection in Remote Sensing Images[J].IEEE Transactions on Geoscience and Remote Sensing,2022,60:1-20.
[15]LU G Q,ZHAO X S,YIN J,et al.Multi-task learning using va-riational auto-encoder for sentiment classification[J].Pattern Recognition Letters,2020,132:115-122.
[16]LU G Q,GAN J Z,YIN J,et al.Multi-task learning using a hybrid representation for text classification[J].Neural Computing and Applications,2020,32(11):6467-6480.
[17]BANIAT L H,PARK S,PARK S B.A Neural Machine Translation Model for Arabic Dialects That Utilizes Multitask Lear-ning(MTL)[J].Computational Intelligence and Neuroscience,2018,2018:1-11.
[18]JI J L,CHEN X H,LUO C Q,et al.A Deep Multi-task Learning Approach for ECG Data Analysis[C]//Proceedings of the 2018 IEEE EMBS International Conference on Biomedical and Health Informatics.Las Vegas,USA,2018:124-127.
[19]LIU Y,LI Q C,WANG K Q,et al.Automatic Multi-Label ECG Classification with Category Imbalance and Cost-Sensitive Thresholding[J].Biosensors-Basel,2021,11(11):1-17.
[20]LIU J H,ZENG M,SHAN K,et al.Loss Optimization Based Algorithm for Multi-classification of ECG Signal[C]//ISAIMS 2022:2022 3rd International Symposium on Artificial Intelligence for Medicine Sciences.2022:537-541.
[21]UBEYLI E D.ECG beats classification using multiclass support vector machines with error correcting output codes[J].Digital Signal Processing,2007,17(3):675-684.
[22]OBERMEYE Z,EMANUEL E J.Predicting the Future－Big Data,Machine Learning,and Clinical Medicine[J].New England Journal of Medicine,2016,375(13):1216-1219.
[23]GUO M,HAQUE A,HUANG D A,et al.Dynamic Task Prioritization for Multitask Learning[C]//Proceedings of the 15th European Conference on Computer Vision.2018:282-299.
[24]YU T H,KUMAR S,GUPT A,et al.Gradient surgery formulti-task learning[C]//Proceedings of the 34th Conference on Neural Information Processing Systems.2020:1-27.
[25]WAGNER P,STRODTHOFF N,BOUSSELJO R-D,et al.PTB-XL,a large publicly available electrocardiography dataset[J].Scientific Data,2020,7(1):1-15.
[26]STRODTHOFF N,WAGNER P,et al.Deep Learning for ECG Analysis:Benchmarks and Insights from PTB-XL[J].IEEE Journal Biomedical and Health Informatics,2021,25(5):1519-1528.
[27]LOSHCHILO I,HUTTER F.Fixing Weight Decay Regularization in ADAM[C]//Proceedings of the 2018 International Conference on Learning Representations.2018:1-10.
[28]CHEN Z,BADRINARAYANA V,LEE C Y,et al.GradNorm:Gradient Normalization for Adaptive Loss Balancing in Deep Multitask Networks[C]//Proceedings of the 35th International Conference on Machine Learning.Stockholm,Sweden,2018:1-27.
[29]LIU S K,JOHNS E,DAVISON A J.End-To-End Multi-Task Learning with Attention[C]//Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition.Long Beach,USA,2019:1871-1880.
[30]CENG Q C,LIU H,GAO T L,et al.An ECG Classification Method Based on Multi-Task Learning and CoT Attention Mechanism[J].Healthcare,2023,11(7):1-13.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发