融合CNN和BiLSTM的心律失常心拍分类模型

doi:10.3969/j.issn.0258-8021.2020.06.009

摘要
图/表
参考文献
相关文章 (15)

全文: PDF (2718 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要为更加准确地从动态心电中提取异常心拍,设计一种融合卷积神经网络(CNN)和多层双边长短时记忆网络(BiLSTM)的心律失常心拍分类模型。心电信号首先被分割成0.75 s和4 s两种不同尺度大小的心拍信号,然后利用11层CNN网络和3层BiLSTM网络分别对小/大尺度心拍信号进行特征提取与合并,并使用3层全连接网络对合并特征进行降维,最后利用softmax函数实现分类。针对MIT心律失常数据库异常心拍类型分布不均衡的问题,采用添加随机运动噪声和基线漂移噪声的样本扩展方法,降低模型的过拟合。采用基于患者的5折交叉检验进行模型验证。MIT心律失常数据库116 000个心拍的分类结果表明:所建立的模型针对4类心拍(正常、房性早搏、室性早搏、未分类)的识别准确率为90.42%,比单独使用CNN(76.45%)和BiLSTM(83.28%)的模型分别提高13.97%和7.14%。所提出的融合CNN和BiLSTM的心律失常心拍分类模型,相比单一基于CNN模型或者BiLSTM模型的机器学习算法,有更好的异常心拍分类准确率。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	杨浩
	黄茂林
	蔡志鹏
	姚映佳
	李建清
	刘澄玉

关键词 ：心律失常, 心拍分类, 心电, 卷积神经网络, 双边长短时记忆网络

Abstract：In order to extract the abnormal beats more accurately from the dynamic electrocardiograph (ECG), a deep learning model combining convolutional neural network (CNN) and bi-directional long short-term memory network (BiLSTM) was proposed in this study. Firstly, ECG signals were segmented into two types of time window lengths: a small-scale length of 0.75 s and a large-scale length of 4 s. Then, features were extracted from the small- and large-scale length ECG segments using an 11-layer CNN network and a 3-layer BiLSTM network, respectively. Finally, the extracted features were combined and were then reduced using a 3-layer fully connected network. In addition, two data enhancement methods by adding random motion noise and baseline drift were used to attenuate the influence of over-fitting due to the unbalanced data distribution. The proposed model was tested on the MIT arrhythmia database using a patient-based 5-fold cross-validation method, and its accuracy for classifying the 4 types (normal, atrial premature, ventricular premature and unclassified) on 116,000 heartbeats was 90.42%, which was 13.97% and 7.14% higher than the CNN model (76.45%) and BiLSTM model (83.28%), respectively. This study validated that the proposed model with combining CNN and BiLSTM reports higher accuracy than only using CNN or BiLSTM model when performing the abnormal beat classification task.

Key words： arrhythmia beat classification electrocardiograms (ECG) convolutional neural network (CNN) bi-directional long short-term memory network (BiLSTM)

收稿日期: 2019-03-11

PACS:

R318

基金资助:国家自然科学基金(81871444);江苏省杰出青年基金(BK20190014);江苏省重点研发计划项目(BE2017735)

通讯作者: E-mail: chengyu@seu.edu.cn

引用本文:

杨浩, 黄茂林, 蔡志鹏, 姚映佳, 李建清, 刘澄玉. 融合CNN和BiLSTM的心律失常心拍分类模型[J]. 中国生物医学工程学报, 2020, 39(6): 719-726.
Yang Hao, Huang Maolin, Cai Zhipeng, Yao Yingjia, Li Jianqing, Liu Chengyu. Arrhythmia Beat Classification Model Based on CNN and BiLSTM. Chinese Journal of Biomedical Engineering, 2020, 39(6): 719-726.

链接本文:

http://cjbme.csbme.org/CN/10.3969/j.issn.0258-8021.2020.06.009 或 http://cjbme.csbme.org/CN/Y2020/V39/I6/719

[1] 上海市肾脏病心血管并发症调查协作组. 上海地区慢性肾功能衰竭患者心血管并发症的调查[J]. 中华肾脏病杂志, 2001, 34(2): 33-38.
[2] 戴闺柱. 心力衰竭诊断与治疗研究进展[J]. 中华心血管病杂志, 2003, 55(9): 9-11.
[3] 刘澄玉, 杨美程, 邸佳楠, 等. 穿戴式心电:发展历程、核心技术与未来挑战[J]. 中国生物医学工程学报, 2019, 38(6): 641-652.
[4] Pan J, Tompkins WJ. A real-Time QRS detection algorithm [J]. IEEE Transactions on Biomedical Engineering, 2007, 2(3): 230-236.
[5] Krasteva V, Jekova I, Leber R, et al. Superiority of classification tree versus cluster, fuzzy and discriminant models in a heartbeat classification System [J]. PLoS ONE, 2015, 10(10): 123-140.
[6] Kedir-Talha MD, Ould-Slimane S. Neural networks and SVM for heartbeat classification[C]//2012 11th International Conference on Information Science, Signal Processing and their Applications (ISSPA). Berlin: IEEE, 2012: 830-835.
[7] Park J, Kang K. PcHD: Personalized classification of heartbeat types using a decision tree[J]. Computers in Biology & Medicine, 2014, 54(1): 79-88.
[8] Yin Wenfeng, Yang Xiuzhu, Lin Zhang, et al. ECG monitoring system integrated with IR-UWB radar based on CNN [J]. IEEE Access, 2016, 4(99): 6344-6351.
[9] Latif S, Usman M, Rana R, et al. Phonocardiographic sensing using deep learning for Abnormal heartbeat detection [J]. IEEE Sensors Journal, 2018, 18(22): 9393-9400.
[10] Sharif Razavian A, Azizpour H, Sullivan J, et al. CNN features off-the-shelf: an astounding baseline for recognition[C]//Proceedings of the IEEE Conference on Computer Vision And Pattern Recognition Workshops. Chicago: IEEE, 2014: 806-813.
[11] Zhao Rui, Yan Ruqiang, Wang Jinjiang, et al. Learning to monitor machine health with convolutional Bi-Directional LSTM Networks[J]. Sensors, 2017, 17(2): 273-278.
[12] Zeng Chao, Cui Songye Xu Xiaowen. Research on resampling of ECG data from MIT-BIH database[J]. Computer Engineering & Applications, 2011, 16(5): 13-17.
[13] Shen Yufei, Zhao Yan. A signal generator using standard MIT-BIH arrhythmia database[J]. Chinese Journal of Medical Instrumentation, 2007, 31(6): 415-418.
[14] Acharya UR, Fujita H, Lih OS, et al. Application of deep convolutional neural network for automated detection of myocardial infarction using ECG signals[J]. Information Sciences, 2017, 415(1): 190-198.
[15] Greff K, Srivastava RK, Koutník J, et al. LSTM: A search space odyssey [J]. IEEE Transactions on Neural Networks & Learning Systems, 2017, 28(10): 2222-2232.
[16] Ari S, Hembram K, Saha G. Detection of cardiac abnormality from PCG signal using LMS based least square SVM classifier [J]. Expert Systems with Applications, 2010, 37(12): 8019-8026.
[17] 倪志伟. BP网络中激活函数的深入研究[J]. 安徽大学学报(自然科学版), 1997, 39(3): 48-51.
[18] Chen Liangjun, Qu Hua, Zhao Jihong, et al. Efficient and robust deep learning with Correntropy-induced loss function [J]. Neural Computing & Applications, 2016, 27(4): 1019-1031.
[19] 龚安, 张敏. BP网络自适应学习率研究[J]. 计算机科学, 2006, 36(1): 64-68.
[20] 覃光华, 丁晶, 陈彬兵. 预防过拟合现象的人工神经网络训练策略及其应用[J]. 长江科学院院报, 2002, 19(3): 59-61.
[21] 李俭川, 秦国军, 温熙森, 等. 神经网络学习算法的过拟合问题及解决方法[J]. 振动, 测试与诊断, 2002, 22(4): 260-264.
[22] Srivastava N, Hinton G, Krizhevsky A, et al. Dropout: A simple way to prevent neural networks from overfitting [J]. Journal of Machine Learning Research, 2014, 15(1): 1929-1958.
[23] 冯晓宇, 史铁梅, 王鑫璐, 等. 彩色多普勒超声联合弹性超声检查在宫颈占位性病变中的应用价值[J]. 现代肿瘤医学, 2017, 25(20): 3313-3316.
[24] 孙亚楠, 吕可嘉, 张瑞. 一种新的心电信号 R 峰自动检测方法[J]. 西北大学学报自然科学版, 2018, 48(1): 16-23.
[25] De Chazal Philip. Detection of supraventricular and ventricular ectopic beats using a single lead ECG[C]//2013 35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). Osaka:IEEE, 2013: 45-48.
[26] Liu Shinghong, Cheng Dachuan, Lin Chihming. Arrhythmia identification with two-lead electrocardiograms using artificial neural networks and support vector machines for a portable ECG monitor system[J]. Sensors, 2013, 13(1): 813-828.
[27] Li Dan, Zhang Jianxin, Zhang Qiang, et al. Classification of ECG signals based on 1D convolution neural network[C]//19th International Conference on e-Health Networking, Applications and Services. Healthcom: IEEE, 2017: 1-6.
[28] 杨朔, 蒲宝明, 李相泽, 等. 基于深度长短时记忆神经网络模型的心律失常检测算法[J]. 计算机应用, 2019, 39(3): 930-934.
[29] 张异凡, 黄亦翔, 汪开正, 等. 用于心律失常识别的 LSTM 和 CNN 并行组合模型[J]. 哈尔滨工业大学学报, 2019,51(10): 11-14.
[30] Jun TJ, Nguyen HM, Kang D, et al. ECG arrhythmia classification using a 2-D convolutional neural network[J]. Scientific Reports, 2018(19): 24-29.
[31] Qin Qin, Li Jianqing, Zhang Li, et al. Combining low-dimensional wavelet features and support vector machine for arrhythmia beat classification[J]. Scientific Reports, 2017, 7(1): 1-12.
[32] Kampouraki A, Manis G, Nikou C. Heartbeat time series classification with support vector machines[J]. Medicine and Biology Society, 2008, 13(4): 512-518.
[33] Chen Taihai, Mazomenos EB, Maharatna K, et al. Design of a low-power on-body ECG classifier for remote cardiovascular monitoring systems[J]. IEEE Journal on Emerging and Selected Topics in Circuits and Systems, 2013, 3(1): 75-85.