正在加载图片...
工程科学学报.第43卷.第9期:1224-1232.2021年9月 Chinese Journal of Engineering,Vol.43,No.9:1224-1232,September 2021 https://doi.org/10.13374/j.issn2095-9389.2021.01.12.005;http://cje.ustb.edu.cn 基于一维卷积特征与手工特征融合的集成超限学习机 心跳分类方法 许越凡12),肖文栋12,3),曹征涛4区 1)北京科技大学自动化学院,北京1000832)北京市工业波谱成像工程技术研究中心,北京1000833)北京科技大学顺德研究生院,顺 德5283994)空军特色医学中心,北京100142 ☒通信作者,E-mail:czhengtao@126.com 摘要融合手工特征和深度特征,提出了一种集成超限学习机心跳分类方法.手工提取的特征明确地表征了心电信号的特 定特性,如相邻心跳时间间隔反映了心跳信号的时域特性,小波系数反映了心跳信号的时频特性.同时设计了一维卷积神经 网络对心跳信号特征进行自动提取.基于超限学习机(Extreme leaning machine,ELM),将上述特征融合进行心跳分类.,由于 ELM初始参数的随机给定可能导致其性能不稳定,进一步提出了一种基于袋装(Bagging)策略的多个ELM集成方法,使分类 结果更加稳定且模型泛化能力更强.利用麻省理工心律失常公开数据集对所提方法进行了验证,分类准确率达到了99.02%, 实验结果也表明基于融合特征的分类准确率高于基于单独特征的分类准确率. 关键词心跳分类:特征融合:一维卷积神经网络:小波变换:集成超限学习机 分类号TP182 Ensemble extreme learning machine approach for heartbeat classification by fusing 1d convolutional and handcrafted features XU Yue-fan2).XIAO Wen-dong2),CAO Zheng-lao 1)School of Automation Electrical Engineering,University of Science and Technology Beijing,Beijing 100083,China 2)Beijing Engineering Research Center of Industrial Spectrum Imaging,Beijing 100083,China 3)Shunde Graduate School,University of Science and Technology Beijing,Shunde 528399,China 4)Air Force Medical Center,PLA,Beijing 100142,China Corresponding author,E-mail:czhengtao@126.com ABSTRACT Arrhythmia is a common cardiovascular disease whose occurrence is mainly related to two factors:cardiac pacing and conduction.Some severe arrhythmias can even threaten human life.An electrocardiogram (ECG)records the changes in electrical activity generated during each cardiac cycle of the heart,which can reflect the human cardiac health status and help diagnose arrhythmias.However,because of the brevity of conventional ECGs,arrhythmias,which occasionally occur in daily life,cannot be detected easily.Automatic ECG analysis-based long-term heartbeat monitoring is of great significance for the effective detection of accidental arrhythmias and then for taking indispensable measures to prevent cardiovascular diseases in time.An ensemble extreme learning machine (ELM)approach for heartbeat classification that fuses handcrafted features and deep features was proposed.The manually extracted features clearly characterize the heartbeat signal,where RR interval features reflect the time-domain characteristic, and the wavelet coefficient features reflect the time-frequency characteristic.A 1D convolutional neural network (1D CNN)was 收稿日期:2021-01-12 基金项目:国家重点研发计划课题资助项目(2017YFB1401203):佛山市科技创新专项资金资助项目(BK20AF005)基于一维卷积特征与手工特征融合的集成超限学习机 心跳分类方法 许越凡1,2),肖文栋1,2,3),曹征涛4) 苣 1) 北京科技大学自动化学院,北京 100083    2) 北京市工业波谱成像工程技术研究中心,北京 100083    3) 北京科技大学顺德研究生院,顺 德 528399    4) 空军特色医学中心,北京 100142 苣通信作者,E-mail: czhengtao@126.com 摘    要    融合手工特征和深度特征,提出了一种集成超限学习机心跳分类方法. 手工提取的特征明确地表征了心电信号的特 定特性,如相邻心跳时间间隔反映了心跳信号的时域特性,小波系数反映了心跳信号的时频特性. 同时设计了一维卷积神经 网络对心跳信号特征进行自动提取. 基于超限学习机(Extreme leaning machine,ELM),将上述特征融合进行心跳分类. 由于 ELM 初始参数的随机给定可能导致其性能不稳定,进一步提出了一种基于袋装(Bagging)策略的多个 ELM 集成方法,使分类 结果更加稳定且模型泛化能力更强. 利用麻省理工心律失常公开数据集对所提方法进行了验证,分类准确率达到了 99.02%, 实验结果也表明基于融合特征的分类准确率高于基于单独特征的分类准确率. 关键词    心跳分类;特征融合;一维卷积神经网络;小波变换;集成超限学习机 分类号    TP182 Ensemble extreme learning machine approach for heartbeat classification by fusing 1d convolutional and handcrafted features XU Yue-fan1,2) ,XIAO Wen-dong1,2,3) ,CAO Zheng-tao4) 苣 1) School of Automation & Electrical Engineering, University of Science and Technology Beijing, Beijing 100083, China 2) Beijing Engineering Research Center of Industrial Spectrum Imaging, Beijing 100083, China 3) Shunde Graduate School, University of Science and Technology Beijing, Shunde 528399, China 4) Air Force Medical Center, PLA, Beijing 100142, China 苣 Corresponding author, E-mail: czhengtao@126.com ABSTRACT    Arrhythmia is a common cardiovascular disease whose occurrence is mainly related to two factors: cardiac pacing and conduction.  Some  severe  arrhythmias  can  even  threaten  human  life.  An  electrocardiogram  (ECG)  records  the  changes  in  electrical activity  generated  during  each  cardiac  cycle  of  the  heart,  which  can  reflect  the  human  cardiac  health  status  and  help  diagnose arrhythmias.  However,  because  of  the  brevity  of  conventional  ECGs,  arrhythmias,  which  occasionally  occur  in  daily  life,  cannot  be detected  easily.  Automatic  ECG  analysis-based  long-term  heartbeat  monitoring  is  of  great  significance  for  the  effective  detection  of accidental  arrhythmias  and  then  for  taking  indispensable  measures  to  prevent  cardiovascular  diseases  in  time.  An  ensemble  extreme learning  machine  (ELM)  approach  for  heartbeat  classification  that  fuses  handcrafted  features  and  deep  features  was  proposed.  The manually extracted features clearly characterize the heartbeat signal, where RR interval features reflect the time-domain characteristic, and  the  wavelet  coefficient  features  reflect  the  time –frequency  characteristic.  A  1D  convolutional  neural  network  (1D  CNN)  was 收稿日期: 2021−01−12 基金项目: 国家重点研发计划课题资助项目(2017YFB1401203);佛山市科技创新专项资金资助项目(BK20AF005) 工程科学学报,第 43 卷,第 9 期:1224−1232,2021 年 9 月 Chinese Journal of Engineering, Vol. 43, No. 9: 1224−1232, September 2021 https://doi.org/10.13374/j.issn2095-9389.2021.01.12.005; http://cje.ustb.edu.cn
<<向上翻页向下翻页>>
©2008-现在 cucdc.com 高等教育资讯网 版权所有