Multi-class 12-lead ECG automatic diagnosis based on a novel subdomain adaptive deep network
Multi-class 12-lead ECG automatic diagnosis based on a novel subdomain adaptive deep network作者机构:State Key Laboratory of Mechanical System and VibrationSchool of Mechanical EngineeringShanghai Jiao Tong UniversityShanghai 200240China Department of Cardiology Shanghai First People's Hospital Afiliated to Shanghai Jiao Tong UniversityShanghai 200080China
出 版 物:《Science China(Technological Sciences)》 (中国科学(技术科学英文版))
年 卷 期:2022年第65卷第11期
页 面:2617-2630页
核心收录:
学科分类:0831[工学-生物医学工程(可授工学、理学、医学学位)] 0711[理学-系统科学] 07[理学] 08[工学] 1010[医学-医学技术(可授医学、理学学位)] 080401[工学-精密仪器及机械] 0804[工学-仪器科学与技术] 080402[工学-测试计量技术及仪器] 10[医学]
基 金:supported by the National Key R&D Program of China(Grant No.2018YFB1307005) the State Key Laboratory of Mechanical System and Vibration(Grant No.MSVZD202103) Shanghai Municipal Science and Technology Major Project(Grant No.2021SHZDZX0102)
主 题:arrhythmia detection subdomainadaptation deep network 12-lead ECG
摘 要:Arrhythmia is a common type of cardiovascular disease,which has become the leading cause of global ***,the automatic 12-lead ECG diagnosis system based on numerous labelled data has attracted increasing ***,labelling 12-lead ECG recordings is a complex and time-consuming task for *** then,the existence of data distribution differences limits the direct cross domain use of the trained *** by subdomain adaptation methods,this paper designs a novel subdomain adaptative deep network(SADN)for excavating diagnosis knowledge from source domain ***,the convolutional layer,residual blocks and SE-Residual blocks are utilized for extracting meaningful deep features ***,the feature classifier uses these deep features for obtaining the final diagnosis ***,designing a novel loss function with local maximum mean discrepancy is utilized for restricting data distribution discrepancy from different ***,the Clinical Outcomes in Digital ECG and 1st China Physiological Signal Challenge datasets are utilized for evaluating the superiority of SADN,which presents that SADN enhances algorithm performance on the unlabelled target domain ***,compared with the existing methods,the proposed network structure acquires better performance with a F1-macro of 89.43%and a F1-macro1 of 87.09%.Besides,among the 4 kinds of ECG abnormalities,the diagnostic effect of the SADN is better than that of cardiology ***,SADN has promising potential as an auxiliary diagnostic tool for the clinical environment.