基于深度学习的心电图心律失常分类方法  

ECG arrhythmia classification method based on deep learning

作　　者：张逸 周莉[1] 陈杰[1] ZHANG Yi;ZHOU Li;CHEN Jie(Institute of Microelectronics of the Chinses Academy of Sciences,Beijing 100029,China;University of Chinese Academy of Sciences,Beijing 100049,China)

机构地区：[1]中国科学院微电子研究所,北京100029 [2]中国科学院大学,北京100049

出　　处：《电子设计工程》2022年第7期6-9,14,共5页Electronic Design Engineering

基　　金：国家自然科学基金项目(U1832217)。

摘　　要：近些年来因心血管疾病导致的人类死亡人数不断增加,心律失常是心血管疾病发病前的常见症状。为了提高心电图对心律失常分类的效率和准确率,使医生能对心律失常及时地作出诊断和治疗,提出一种基于二维卷积神经网络模型的心律失常分类方法。该方法使用美国麻省理工学院提供的研究心律失常的MIT-BIH数据库来生成实验数据集对网络进行训练和测试,在心律失常分类测试中分类准确率达到了98.6%,实现了对心电图信号心律失常的高精度自动分类。In recent years,the number of human deaths caused by cardiovascular diseases has been increasing,and arrhythmia is a common symptom before the onset of cardiovascular diseases.In order to improve the efficiency and accuracy of classifying arrhythmia through electrocardiogram,so that doctors can diagnose and treat the arrhythmia in a timely manner,a method is proposed for arrhythmia classification based on a two-dimensional convolutional neural network model,using MIT-BIH database for arrhythmia research provided by the Massachusetts Institute of Technology to generate experimental data sets to train and test the network.In the arrhythmia classification test,it achieved a classification accuracy of 98.6%,and realized high-precision automatic detection of arrhythmia from ECG signals classification.

关 键 词：心律失常 心电图信号 深度学习 卷积神经网络 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]