|本期目录/Table of Contents|

[1]徐晨华,叶思超,乔清理△.基于卷积神经网络特征提取与融合的心律失常分类*[J].生物医学工程研究,2021,01:15-20.
 XU Chenhua,YE Sichao,QIAO Qingli.Arrhythmia classification based on feature extraction and fusion of convolutional neural network[J].Journal of Biomedical Engineering Research,2021,01:15-20.
点击复制

基于卷积神经网络特征提取与融合的心律失常分类*(PDF)

《生物医学工程研究》[ISSN:1006-6977/CN:61-1281/TN]

期数:
2021年01期
页码:
15-20
栏目:
出版日期:
2021-03-25

文章信息/Info

Title:
Arrhythmia classification based on feature extraction and fusion of convolutional neural network
文章编号:
1672-6278 (2021)01-0015-06
作者:
徐晨华叶思超乔清理△
天津医科大学生物医学工程与技术学院,天津 300070
Author(s):
XU ChenhuaYE SichaoQIAO Qingli
School of Biomedical Engineering and Technology,Tianjin Medical University,Tianjin 300070, China
关键词:
心律失常单导联特征提取与融合分类卷积神经网络
Keywords:
Arrhythmia Single lead Feature extraction and fusionClassification Convolutional neural network
分类号:
R318
DOI:
10.19529/j.cnki.1672-6278.2021.01.03
文献标识码:
A
摘要:
本研究提出一种新的心律失常自动分类方法,辅助医生诊治心律失常。通过构建卷积神经网络对心电信号以及QRS波群的小波分量进行特征提取,将网络提取到的心电信号特征和小波特征与人工提取的RR间期特征,输入到全连接层进行融合,在输出层使用softmax函数对心拍进行分类。使用MIT-BIH心律失常数据库中的MILL导联数据对网络进行训练和测试。经测试,该方法的总体分类准确度达98.12%,平均灵敏度为87.32%,平均阳性预测值为90.37%。该方法能够快速识别不同类型的心律失常,对于计算机辅助诊断心律失常的应用具有一定的参考价值。
Abstract:
We proposed a new automatic classification method of arrhythmia to help doctors prevent and diagnose arrhythmia. The convolution neural network was constructed to extract the features of electrocardiogram(ECG) and QRS complex wavelet. The ECG signal features, wavelet features and the manually extracted RR interval features were inputted to the fully connected layer for fusion. The heartbeats were classified by softmax function in the output layer. The network was trained and tested using the data of MILL lead in MIT-BIH arrhythmia database. The results showed that the total classification accuracy was 98.12%, the average sensitivity was 87.32% and the average positive predictive value was 90.37%. This method can quickly identify different types of arrhythmia, and has certain reference value for the application of computer-aided diagnosis of arrhythmia.

参考文献/References

备注/Memo

备注/Memo:
(收稿日期:2020-07-17)国家自然科学基金资助项目(30870649)。△通信作者Email:qlqiao@tmu.edu.cn
更新日期/Last Update: 2021-04-13