[1]查雪帆,杨丰,吴俣南,等. 结合迁移学习与深度卷积网络的心电分类研究[J].中国医学物理学杂志,2018,35(11):1307-1312.[doi:DOI:10.3969/j.issn.1005-202X.2018.11.013]
 ZHA Xuefan,YANG Feng,WU Yunan,et al. ECG classification based on transfer learning and deep convolution neural network[J].Chinese Journal of Medical Physics,2018,35(11):1307-1312.[doi:DOI:10.3969/j.issn.1005-202X.2018.11.013]
点击复制

 结合迁移学习与深度卷积网络的心电分类研究()
分享到:

《中国医学物理学杂志》[ISSN:1005-202X/CN:44-1351/R]

卷:
35卷
期数:
2018年第11期
页码:
1307-1312
栏目:
医学信号处理与医学仪器
出版日期:
2018-11-18

文章信息/Info

Title:
 ECG classification based on transfer learning and deep convolution neural network
文章编号:
1005-202X(2018)11-1307-06
作者:
 查雪帆1杨丰12吴俣南1刘颖1袁绍锋12
 1.南方医科大学生物医学工程学院, 广东 广州 510515; 2.南方医科大学广东省医学图像处理重点实验室, 广东 广州 510515
Author(s):
 ZHA Xuefan1 YANG Feng1 2 WU Yu’nan1 LIU Ying1 YUAN Shaofeng1 2
 1. School of Biomedical Engineering, Southern Medical University, Guangzhou 510515, China; 2. Guangdong Provincial Key Laboratory of Medical Image Processing, Southern Medical University, Guangzhou 510515, China
关键词:
心电节拍分类迁移学习深度学习二维深度卷积网络一维深度卷积网络ImageNet数据集
Keywords:
 Keywords: electrocardiogram heartbeat classification transfer learning deep learning two-dimensional deep convolution neural network one-dimensional deep convolution neural network ImageNet dataset
分类号:
R318
DOI:
DOI:10.3969/j.issn.1005-202X.2018.11.013
文献标志码:
A
摘要:
 为解决一维深度卷积网络(1D-DCNN)在心电分类方面存在的多类疾病识别不准、难以提取最佳特征等问题,提出一种结合迁移学习与二维深度卷积网络(2D-DCNN)直接识别心电图像的方法。首先,截取R波前后75 ms内的心电信号,并将一维心电电压信号转化为二维灰度图像信号。接着,构建2D-DCNN对心电节拍样本进行分类训练,权值初始化采用在ImageNet大规模图像数据集上进行预训练的AlexNet参数值。本文提出方法在MIT-BIH心电数据库上进行性能验证,其准确率达到98%,并在不同信噪比下保持较高的准确率,证明了所述模型在心电分类上具有良好的鲁棒性。为了验证2D-DCNN的识别性能,实验部分与采用不同激活函数的1D-DCNN、近些年性能较好的深度学习方法进行比较。量化结果表明,结合迁移学习和2D-DCNN方法,比最优1D-DCNN算法,其准确率提升2%、敏感度提升0.6%、特异性提高4%;在二分类与多分类任务中,均好于现有的其他算法。
Abstract:
 Abstract: One-dimensional deep convolution neural networks (1D-DCNN) for electrocardiogram (ECG) classification shows limitations in identifying various diseases and extracting the best morphological features. Herein, a method combining transfer learning and two-dimensional deep convolution neural network (2D-DCNN), AlexNet, is proposed to identify ECG images directly. Firstly, the ECG signals within the 75 ms before and after R wave were intercepted, and one-dimensional ECG voltage signals were converted into two-dimensional grayscale image signals. Then, a 2D-DCNN based on AlexNet was established to classify the ECG heartbeat samples. The weights were initialized by parameters which were pre-trained on Alexnet using a large-scale dataset ImageNet. The proposed method achieved an accuracy of 98% on the MIT-BIH arrhythmia database, and maintained a high accuracy at different signal-to-noise ratios, which verified the good robustness of the proposed method in ECG classification. The proposed method was also compared with 1D-DCNN using different activation functions and other deep learning methods with favorable performances to evaluate the performance of 2D-DCNN. The quantitative results demonstrated that compared with the optimal 1D-DCNN, the proposed method combining transfer learning with 2D-DCNN improves the accuracy rate, sensitivity and specificity by 2%, 0.6% and 4%, respectively, and that the proposed algorithm is better than other existing algorithms in both binary/multi-class classification tasks.

相似文献/References:

[1]芮迎迎,孔祥勇,刘亚楠,等.基于Mask Scoring R-CNN的齿痕舌象识别[J].中国医学物理学杂志,2021,38(4):523.[doi:DOI:10.3969/j.issn.1005-202X.2021.04.023]
 RUI Yingying,KONG Xiangyong,LIU Yanan,et al.Tooth-marked tongue recognition using Mask Scoring R-CNN[J].Chinese Journal of Medical Physics,2021,38(11):523.[doi:DOI:10.3969/j.issn.1005-202X.2021.04.023]
[2]吴雪,王娆芬.基于迁移学习的深层卷积神经网络心电信号疲劳分类[J].中国医学物理学杂志,2021,38(10):1258.[doi:DOI:10.3969/j.issn.1005-202X.2021.10.013]
 WU Xue,WANG Raofen.Classification of fatigue state using ECG signals and DCNN with transfer learning[J].Chinese Journal of Medical Physics,2021,38(11):1258.[doi:DOI:10.3969/j.issn.1005-202X.2021.10.013]
[3]赵清一,林勇.基于迁移学习和支持向量机的乳腺癌分子分型预测[J].中国医学物理学杂志,2022,39(5):635.[doi:DOI:10.3969/j.issn.1005-202X.2022.05.019]
 ZHAO Qingyi,LIN Yong.Breast cancer molecular typing prediction based on transfer learning and support vector machine[J].Chinese Journal of Medical Physics,2022,39(11):635.[doi:DOI:10.3969/j.issn.1005-202X.2022.05.019]
[4]冀常鹏,杨梦晗,代巍.基于改进Faster RCNN的舌部多纹理检测[J].中国医学物理学杂志,2023,40(8):977.[doi:DOI:10.3969/j.issn.1005-202X.2023.08.009]
 JI Changpeng,YANG Menghan,DAI Wei.Tongue multi-texture recognition using improved Faster RCNN[J].Chinese Journal of Medical Physics,2023,40(11):977.[doi:DOI:10.3969/j.issn.1005-202X.2023.08.009]
[5]吴书裕,周露,王琳婧,等.基于BIRADs多任务学习模型的乳腺肿块分类[J].中国医学物理学杂志,2023,40(10):1220.[doi:DOI:10.3969/j.issn.1005-202X.2023.10.005]
 WU Shuyu,ZHOU Lu,WANG Linjing,et al.Breast mass classification based on BIRADs multi-task learning model[J].Chinese Journal of Medical Physics,2023,40(11):1220.[doi:DOI:10.3969/j.issn.1005-202X.2023.10.005]

备注/Memo

备注/Memo:
 【收稿日期】2018-06-28
【基金项目】国家自然科学基金(61771233, 61271155)
【作者简介】查雪帆,研究方向:机器学习与医学图像处理,E-mail: xuefanzha.smu@gmail.com
【通信作者】杨丰,教授,博士生导师,研究方向:模式识别、机器学习、医学图像处理、医学信号处理,E-mail: yangf@smu.edu.cn
更新日期/Last Update: 2018-11-22