高级搜索

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

基于自适应快速S变换和XGBoost的心电信号精确快速分类方法

袁莉芬 李松 尹柏强 李兵 佐磊

袁莉芬, 李松, 尹柏强, 李兵, 佐磊. 基于自适应快速S变换和XGBoost的心电信号精确快速分类方法[J]. 电子与信息学报, 2023, 45(4): 1464-1474. doi: 10.11999/JEIT220217
引用本文: 袁莉芬, 李松, 尹柏强, 李兵, 佐磊. 基于自适应快速S变换和XGBoost的心电信号精确快速分类方法[J]. 电子与信息学报, 2023, 45(4): 1464-1474. doi: 10.11999/JEIT220217
YUAN Lifen, LI Song, YIN Baiqiang, LI Bing, ZUO Lei. Accurate and Fast ElectroCardioGram Classification Method Based on Adaptive Fast S-Transform and XGBoost[J]. Journal of Electronics & Information Technology, 2023, 45(4): 1464-1474. doi: 10.11999/JEIT220217
Citation: YUAN Lifen, LI Song, YIN Baiqiang, LI Bing, ZUO Lei. Accurate and Fast ElectroCardioGram Classification Method Based on Adaptive Fast S-Transform and XGBoost[J]. Journal of Electronics & Information Technology, 2023, 45(4): 1464-1474. doi: 10.11999/JEIT220217

基于自适应快速S变换和XGBoost的心电信号精确快速分类方法

doi: 10.11999/JEIT220217
基金项目: 国家自然科学基金(61971175),中央高校基本科研业务费(JZ2019YYPY0025)
详细信息
    作者简介:

    袁莉芬:女,博士,教授,研究方向为射频识别技术

    李松:男,硕士生,研究方向为信号处理与分析

    尹柏强:男,博士,教授,研究方向为电能质量先进检测与控制方法

    李兵:男,博士,教授,研究方向为智能电网信息工程

    佐磊:男,博士,副研究员,研究方向为智能感知技术及应用

    通讯作者:

    尹柏强 yinbaiqiang123@163.com

  • 中图分类号: TN911.7; R540.41

Accurate and Fast ElectroCardioGram Classification Method Based on Adaptive Fast S-Transform and XGBoost

Funds: The National Natural Science Foundation of China (61971175), The Fundamental Research Funds for the Central Universities (JZ2019YYPY0025)
  • 摘要: 针对心电信号(ECG)传统分类方法效率较低的问题,该文提出一种基于自适应快速S变换(AFST)和XGBoost的心电信号精确快速分类方法。该方法首先通过快速定位算法确定心电信号特征频率点,再根据特征频率点自适应调节S变换窗宽因子,增强S变换的时频分辨率的同时避免迭代计算,大大减少运行时间。其次,基于自适应快速S变换的时频矩阵提取12个特征量来表征5种心电信号的特征信息,特征向量维数低,识别能力强。最后,利用XGBoost算法对特征向量进行识别。MIT-BIH心律失常数据库和患者实测数据验证表明,该方法显著地缩短了分类时间,对5种心电信号的分类准确率分别为99.59%和97.32%,适用于实际检测系统中心律失常疾病的快速诊断。
  • 图  1  不同$ \beta (f) $对应Kaiser窗幅频特性

    图  2  心电信号频率包络线

    图  3  心电信号时域累积特性曲线

    图  4  不同样本长度算法运行时间对比

    图  5  心电信号时域波形图

    图  6  心电信号AFST变换时频3维图

    图  7  基于ST提取的特征可视化分布情况

    图  8  基于AFST提取的特征可视化分布情况

    图  9  不同特征提取方法性能比较

    图  10  不同分类器性能比较

    表  1  不同算法运算量分析

    算法运算量
    复数乘法复数加法时间复杂度
    ST$ ({N^2} + {N^3})/2 $$ ({N^3} - {N^2})/2 $$ O({N^3}) $
    FST$ (2{N^2} + {N^2}{\log _2}N)/4 $$ ({N^2}{\log _2}N)/2 $$ O({N^2}\log {}_2N) $
    AFST$ q(2N + N{\log _2}N)/2 $$ qN{\log _2}N $$ O(qN\log {}_2N) $
    下载: 导出CSV

    表  2  分类结果(%)

    心电类型$ {\text{Acc}} $$ {\text{Sen}} $$ {\text{Spe}} $$ {\text{PPV}} $
    NORM99.6899.0699.3098.95
    LBBB99.4398.8399.5899.64
    RBBB99.5299.5499.9099.96
    APC99.7099.3698.8699.43
    PVC99.6199.4899.7199.29
    总计99.5999.2599.4799.45
    下载: 导出CSV

    表  3  各算法性能对比

    特征提取方法分类器特征提取
    时间(s)
    分类器运行
    时间(s)
    准确率
    (%)
    ST随机森林389.106.4284.14
    GBDT4.9880.29
    XGBoost4.3585.95
    改进XGBoost4.3588.53
    AFST随机森林152.076.4798.42
    GBDT4.9897.74
    XGBoost4.3699.06
    改进XGBoost4.3699.59
    下载: 导出CSV

    表  4  不同分类方法对比(%)

    方法分类器ECG特征AccSenSpePPV
    Wavelet[24]E-SVM小波系数、R-R间期、形态特征94.5094.7093.90
    Wavelet[8]RNN-LSTM小波系数、R-R间期、形态特征97.1085.7094.2898.35
    TSFEL[16]RT+SVM统计学特征、时频特征98.2184.2195.9592.81
    XWT[8]SVM小波系数、R-R间期、时域特征98.5099.6998.80
    SST[25]SVMSST系数、R-R间期、形态特征84.7180.4370.75
    ST[11]GA-SVMR-R间期、时频特征、形态特征99.7499.4299.83
    AFST改进XGBoost时频特征99.5999.2599.4799.45
    下载: 导出CSV
  • [1] KREATSOUS C and ANAND S S. The impact of social determinants on cardiovascular disease[J]. Canadian Journal of Cardiology, 2010, 26 Suppl C: 8C–13C.
    [2] World Health Organization. Cardiovascular disease (CVDs)[EB/OL].https://www.who.int/en/news-room/fact-sheets/detail/cardiovascular-diseases-(cvds), 2021.
    [3] AMORIM P, MORAES T, FAZANARO D, et al. Shearlet and contourlet transforms for analysis of electrocardiogram signals[J]. Computer Methods and Programs in Biomedicine, 2018, 161: 125–132. doi: 10.1016/j.cmpb.2018.04.021
    [4] 吴志勇, 丁香乾, 许晓伟, 等. 基于深度学习和模糊C均值的心电信号分类方法[J]. 自动化学报, 2018, 44(10): 1913–1920. doi: 10.16383/j.aas.2018.c170417

    WU Zhiyong, DING Xiangqian, XU Xiaowei, et al. A method for ECG classification using deep learning and fuzzy C-means[J]. Acta Automatica Sinica, 2018, 44(10): 1913–1920. doi: 10.16383/j.aas.2018.c170417
    [5] DE CHAZAL P, O’DWYER M, and REILLY R B. Automatic classification of heartbeats using ECG morphology and heartbeat interval features[J]. IEEE Transactions on Biomedical Engineering, 2004, 51(7): 1196–1206. doi: 10.1109/TBME.2004.827359
    [6] MAR T, ZAUNSEDER S, MARTÍNEZ J P, et al. Optimization of ECG classification by means of feature selection[J]. IEEE Transactions on Biomedical Engineering, 2011, 58(8): 2168–2177. doi: 10.1109/TBME.2011.2113395
    [7] 王金海, 史梦颖, 张兴华. 基于EMD和ApEn特征提取的心律失常分类研究[J]. 仪器仪表学报, 2016, 37(S1): 168–173. doi: 10.19650/j.cnki.cjsi.2016.s1.028

    WANG Jinhai, SHI Mengying, and ZHANG Xinghua. Classification of arrhythmias based on EMD and ApEn feature extraction[J]. Chinese Journal of Scientific Instrument, 2016, 37(S1): 168–173. doi: 10.19650/j.cnki.cjsi.2016.s1.028
    [8] JACOB N and JOSEPH L A. Classification of ECG beats using cross wavelet transform and support vector machines[C]. 2015 IEEE Recent Advances in Intelligent Computational Systems (RAICS), Trivandrum, India, 2015.
    [9] 尹柏强, 邓影, 王署东, 等. 时频广义S变换和VL-MOBP神经网络在人体动作识别中的应用[J]. 电子测量与仪器学报, 2020, 34(11): 1–9. doi: 10.13382/j.jemi.B2003032

    YIN Baiqiang, DENG Ying, WANG Shudong, et al. Application of time-frequency generalized S transform and VL-MOBP neural network in human motion recognition[J]. Journal of Electronic Measurement and Instrumentation, 2020, 34(11): 1–9. doi: 10.13382/j.jemi.B2003032
    [10] DAS M K and ARI S. Electrocardiogram beat classification using S-transform based feature set[J]. Journal of Mechanics in Medicine and Biology, 2014, 14(5): 1450066. doi: 10.1142/S0219519414500663
    [11] YANG Jianyong and YAN Ruqiang. A multidimensional feature extraction and selection method for ECG arrhythmias classification[J]. IEEE Sensors Journal, 2021, 21(13): 14180–14190. doi: 10.1109/JSEN.2020.3047962
    [12] RAJ S and RAY K C. ECG signal analysis using DCT-Based DOST and PSO optimized SVM[J]. IEEE Transactions on Instrumentation and Measurement, 2017, 66(3): 470–478. doi: 10.1109/TIM.2016.2642758
    [13] SINHA N and DAS A. Identification and localization of myocardial infarction based on analysis of ECG signal in cross spectral domain using boosted SVM classifier[J]. IEEE Transactions on Instrumentation and Measurement, 2021, 70: 4007409. doi: 10.1109/TIM.2021.3117663
    [14] HOU Borui, YANG Jianyong, WANG Pu, et al. LSTM-based auto-encoder model for ECG arrhythmias classification[J]. IEEE Transactions on Instrumentation and Measurement, 2020, 69(4): 1232–1240. doi: 10.1109/TIM.2019.2910342
    [15] 柯丽, 王丹妮, 杜强, 等. 基于卷积长短时记忆网络的心律失常分类方法[J]. 电子与信息学报, 2020, 42(8): 1990–1998. doi: 10.11999/JEIT190712

    KE Li, WANG Danni, DU Qiang, et al. Arrhythmia classification based on convolutional long short term memory network[J]. Journal of Electronics &Information Technology, 2020, 42(8): 1990–1998. doi: 10.11999/JEIT190712
    [16] BHATTACHARYYA S, MAJUMDER S, DEBNATH P, et al. Arrhythmic heartbeat classification using ensemble of random forest and support vector machine algorithm[J]. IEEE Transactions on Artificial Intelligence, 2021, 2(3): 260–268. doi: 10.1109/TAI.2021.3083689
    [17] KUNG B H, HU P Y, HUANG C C, et al. An efficient ECG classification system using resource-saving architecture and random forest[J]. IEEE Journal of Biomedical and Health Informatics, 2021, 25(6): 1904–1914. doi: 10.1109/JBHI.2020.3035191
    [18] 王帅, 赵钟瑶, 张翔宇, 等. 穿戴式心电信号质量的三分类评估方法[J]. 中国生物医学工程学报, 2020, 39(5): 550–556. doi: 10.3969/j.issn.0258-8021.2020.05.005

    WANG Shuai, ZHAO Zhongyao, ZHANG Xiangyu, et al. Three-type classification method for wearable ECG signal quality[J]. Chinese Journal of Biomedical Engineering, 2020, 39(5): 550–556. doi: 10.3969/j.issn.0258-8021.2020.05.005
    [19] 喻玲, 符玲, 熊思宇, 等. 基于最优Kaiser窗的动态同步相量测量算法[J]. 电网技术, 2022, 46(3): 1100–1108. doi: 10.13335/j.1000-3673.pst.2021.0602

    YU Ling, FU Ling, XIONG Siyu, et al. A dynamic synchrophasor estimation algorithm based on optimal Kaiser window[J]. Power System Technology, 2022, 46(3): 1100–1108. doi: 10.13335/j.1000-3673.pst.2021.0602
    [20] 尹柏强, 何怡刚, 朱彦卿. 一种广义S变换及模糊SOM网络的电能质量多扰动检测和识别方法[J]. 中国电机工程学报, 2015, 35(4): 866–872. doi: 10.13334/j.0258-8013.pcsee.2015.04.013

    YIN Baiqiang, HE Yigang, and ZHU Yanqing. Detection and classification of power quality multi-disturbances based on generalized s-transform and fuzzy SOM neural network[J]. Proceedings of the CSEE, 2015, 35(4): 866–872. doi: 10.13334/j.0258-8013.pcsee.2015.04.013
    [21] 王奕森, 夏树涛. 集成学习之随机森林算法综述[J]. 信息通信技术, 2018, 12(1): 49–55. doi: 10.3969/j.issn.1674-1285.2018.01.009

    WANG Yisen and XIA Shutao. A survey of random forests algorithms[J]. Information and Communications Technologies, 2018, 12(1): 49–55. doi: 10.3969/j.issn.1674-1285.2018.01.009
    [22] 赵洪山, 闫西慧, 王桂兰, 等. 应用深度自编码网络和XGBoost的风电机组发电机故障诊断[J]. 电力系统自动化, 2019, 43(1): 81–86. doi: 10.7500/AEPS20180708001

    ZHAO Hongshan, YAN Xihui, WANG Guilan, et al. Fault diagnosis of wind turbine generator based on deep autoencoder network and XGBoost[J]. Automation of Electric Power Systems, 2019, 43(1): 81–86. doi: 10.7500/AEPS20180708001
    [23] PAN Jiapu and TOMPKINS W J. A real-time QRS detection algorithm[J]. IEEE Transactions on Biomedical Engineering, 1985, BME-32(3): 230–236. doi: 10.1109/TBME.1985.325532
    [24] CHIA N G, HAU Y W, and JAMALUDIN M N. Robust arrhythmia classifier using wavelet transform and support vector machine classification[C]. The 13th International Colloquium on Signal Processing & its Applications (CSPA), Penang, Malaysia, 2017.
    [25] HERRY C L, FRASCH M, SEELY A J, et al. Heart beat classification from single-lead ECG using the synchrosqueezing transform[J]. Physiological Measurement, 2017, 38(2): 171–187. doi: 10.1088/1361-6579/aa5070
  • 加载中
图(10) / 表(4)
计量
  • 文章访问数:  387
  • HTML全文浏览量:  541
  • PDF下载量:  85
  • 被引次数: 0
出版历程
  • 收稿日期:  2022-03-02
  • 修回日期:  2022-07-10
  • 网络出版日期:  2022-07-19
  • 刊出日期:  2023-04-10

目录

    /

    返回文章
    返回