基于自适应向量机检测睡眠呼吸暂停综合征的最优特征组合筛选
|
摘要
基于自适应向量机监测睡眠呼吸暂停综合征(SAS)时可提取出的特征参数较多,筛选这些特征参数中与SAS相关度较大的组合,可以有效降低算法的计算量,具有重要的实践意义。本文基于V2导联心电信号,首先对ECG信号进行去噪和R波提取,得到心率变异性信号(HRV)和心电呼吸导出信号,并从中提取出时域频域特征共22组,利用特征参数与SAS的相关系数对特征参数筛选后进行支持向量机(SVM)分类。对比22组特征参数与筛选后的15组特征参数分类结果,准确率降低不足0.5%,但计算复杂度大大降低,可作为对临床长时间心电图检测的扩展,减少对专业医护人员的依赖,具有良好的经济性和普及性。
There are so much characteristic parameters can be extracted based on adaptive vector machine for detecting sleep apnea syndrome. It has important significance which is selected the characteristic parameters to reduce the amounts of calculation applied in sleep apnea syndrome. This study adopted the electrocardiogram signals from limb guided lead-Ⅱ and then denoised the signal interference and detected the R-wave to get the heart rate variability data and ECG-derived respiratory data. Analysis these two data that we can obtain the twenty-two features in time domain and frequency domain, moreover, using support vector machines algorithm to classify the sleep apnea syndrome feature parameters. Compared the twenty-two features with optimal fifteen feature parameters we proposed, the amounts of calculation are decrease obviously without decay the classification accuracy. It can be used as an extension of clinical long time electrocardiogram detection because it can reduce the dependence on health care professional. Therefore, it has good economy and popularity.
There are so much characteristic parameters can be extracted based on adaptive vector machine for detecting sleep apnea syndrome. It has important significance which is selected the characteristic parameters to reduce the amounts of calculation applied in sleep apnea syndrome. This study adopted the electrocardiogram signals from limb guided lead-Ⅱ and then denoised the signal interference and detected the R-wave to get the heart rate variability data and ECG-derived respiratory data. Analysis these two data that we can obtain the twenty-two features in time domain and frequency domain, moreover, using support vector machines algorithm to classify the sleep apnea syndrome feature parameters. Compared the twenty-two features with optimal fifteen feature parameters we proposed, the amounts of calculation are decrease obviously without decay the classification accuracy. It can be used as an extension of clinical long time electrocardiogram detection because it can reduce the dependence on health care professional. Therefore, it has good economy and popularity.
引文
[1]Hiestand DM,Britz P,Goldman M,et al.Prevalence of symptoms and risk of sleep apnea in the US population:Results from the national sleep foundation sleep in America[J].Chest,2006,130(3):780-786.
[2]Finkel KJ,Searleman AC,Tymkew H,et al.Prevalence of undiagnosed obstructive sleep apnea among adult surgical patients in an academic medical center[J].Sleep Med,2009,10:753-758.
[3]孙薇.睡眠呼吸暂停综合征的监测与治疗关键算法研究[D].广州:南方医科大学,2014.
[4]余晓敏,涂岳文,黄超,等.基于心电信号的睡眠呼吸暂停综合征检测算法[J].生物医学工程学杂志,2013,30(5):999-1002.
[5]吴宪文.基于单导心电检测睡眠呼吸暂停综合症[D].南京:东南大学,2010.
[6]de Chazal P,Heneghan C,Sheridan E,et al.Automatic classification of sleep apnea epochs using the electrocardiogram[J].Comput Cardiol,2000,27:745-748.
[7]MIT-BIH Arrhythmia Database[EB/OL].(2016-06-16).http://www.physionet.org/physiobank/database/mitdb/
[8]Penzel T,Moody GB,Mark RG,et al.The apnea-ECGdatabase[J].Computers in Cardiology,2000(27):255-258.
[9]Moody GB,Mark RG,Goldberger A,et al.Stimulating rapid research advances via focused competition:the computers in cardiology challenge[J].Computers in Cardiology,2000(27):207-210.
[10]毋斌.基于小波变换的心电信号阈值去噪[J].山西科技,2018,33(1):77-79.
[11]郑敏敏,高小榕,谢海鹤.心电信号小波去噪的改进算法研究[J].中国生物医学工程学报,2017,36(1):114-118.
[12]王春雨.基于小波变换的ECG信号分析研究与硬件实现[D].福建:福州大学,2017.
[13]Hu X,Xiao Z,Zhang N.Removal of baseline wander from ECG signal based on a statistical weighted moving average filter[J].J Zhejiang Univ Sci,2011,12(5):397-403.
[14]张清丽,苏士美,王猛.基于改进小波变换的QRS特征提取算法研究[J].郑州大学学报:理学版,2017,49(4):100-103.
[15]李忠杰.临床心脏不应期检测[J].心电学杂志,2007(1):52-55.
[16]林璨璨,刘梅颜.动态心电图初筛睡眠呼吸暂停综合征进展[J].中华医学杂志,2017,97(1):76-78.
[17]刘双艳.睡眠呼吸暂停病人HRV和PRV的相关性研究[D].济南:山东大学,2017.
[18]Cai H,Chen Y,Han J,et al.Study on feature selection methods for depression detection using three-electrode EEG data[J].Interdiscip Sci,2018,10(3):558-565.
[19]de Chazal P,Heneghan C,Sheridan E,et al.Automated processing of the single-lead electrocardiogram for the detection of obstructive sleep apnoea[J].IEEE Trans Biomed Eng,2003,50(6):686-696.
[20]拓守恒.基于QPSO训练的SVM核函数集成学习研究[J].系统仿真技术,2010,6(3):202-208,240.
[2]Finkel KJ,Searleman AC,Tymkew H,et al.Prevalence of undiagnosed obstructive sleep apnea among adult surgical patients in an academic medical center[J].Sleep Med,2009,10:753-758.
[3]孙薇.睡眠呼吸暂停综合征的监测与治疗关键算法研究[D].广州:南方医科大学,2014.
[4]余晓敏,涂岳文,黄超,等.基于心电信号的睡眠呼吸暂停综合征检测算法[J].生物医学工程学杂志,2013,30(5):999-1002.
[5]吴宪文.基于单导心电检测睡眠呼吸暂停综合症[D].南京:东南大学,2010.
[6]de Chazal P,Heneghan C,Sheridan E,et al.Automatic classification of sleep apnea epochs using the electrocardiogram[J].Comput Cardiol,2000,27:745-748.
[7]MIT-BIH Arrhythmia Database[EB/OL].(2016-06-16).http://www.physionet.org/physiobank/database/mitdb/
[8]Penzel T,Moody GB,Mark RG,et al.The apnea-ECGdatabase[J].Computers in Cardiology,2000(27):255-258.
[9]Moody GB,Mark RG,Goldberger A,et al.Stimulating rapid research advances via focused competition:the computers in cardiology challenge[J].Computers in Cardiology,2000(27):207-210.
[10]毋斌.基于小波变换的心电信号阈值去噪[J].山西科技,2018,33(1):77-79.
[11]郑敏敏,高小榕,谢海鹤.心电信号小波去噪的改进算法研究[J].中国生物医学工程学报,2017,36(1):114-118.
[12]王春雨.基于小波变换的ECG信号分析研究与硬件实现[D].福建:福州大学,2017.
[13]Hu X,Xiao Z,Zhang N.Removal of baseline wander from ECG signal based on a statistical weighted moving average filter[J].J Zhejiang Univ Sci,2011,12(5):397-403.
[14]张清丽,苏士美,王猛.基于改进小波变换的QRS特征提取算法研究[J].郑州大学学报:理学版,2017,49(4):100-103.
[15]李忠杰.临床心脏不应期检测[J].心电学杂志,2007(1):52-55.
[16]林璨璨,刘梅颜.动态心电图初筛睡眠呼吸暂停综合征进展[J].中华医学杂志,2017,97(1):76-78.
[17]刘双艳.睡眠呼吸暂停病人HRV和PRV的相关性研究[D].济南:山东大学,2017.
[18]Cai H,Chen Y,Han J,et al.Study on feature selection methods for depression detection using three-electrode EEG data[J].Interdiscip Sci,2018,10(3):558-565.
[19]de Chazal P,Heneghan C,Sheridan E,et al.Automated processing of the single-lead electrocardiogram for the detection of obstructive sleep apnoea[J].IEEE Trans Biomed Eng,2003,50(6):686-696.
[20]拓守恒.基于QPSO训练的SVM核函数集成学习研究[J].系统仿真技术,2010,6(3):202-208,240.