心电信号特征量的小波提取算法  

Detection of ECG Characteristic Points by Wavelet Transform

作　　者：徐舫舟[1] 赵捷[1] 赵艳娜[1] 田杰[1] 

机构地区：[1]山东师范大学物理与电子科学学院,山东济南250014

出　　处：《现代生物医学进展》2009年第17期3319-3322,共4页Progress in Modern Biomedicine

基　　金：山东省自然科学基金(Y2007Z05);山东省科技攻关项目(2007GG10001018)

摘　　要：目的:为了更准确地利用心电图(ECG)进行临床生理疾病诊断,提高心电信号的自动分析准确度。介绍了一种利用小波变换的时频局部化特性以及多分辨率特性对心电信号进行处理的算法。方法:使用定位准确,计算简便的二阶微分Marr小波使用多孔算法来对ECG中QRS波群进行标定。结果:将算法应用到MIT/BIH国际标准心电数据库进行仿真。结论:通过仿真证明,该算法能够很精确地定位QRS波群,为心电信号的后续研究打好基础。Objective： To better use electrocardiogram （ECG） to complete clinical diagnosis of physical diseases,this paper introduces an algorithm about ECG characteristic points based on wavelet transform, which enjoys time-frequency localization and multi-resolution characters. Method： Based on a-trous algorithm, second order differential Marr wavelet,which has the advantage of accurate positioning and simple calculation, is used to detect QRS complexes from ECG. Results： This algorithm is evaluated on MIT-BIH Arrhythmia database. Conclusion： Through simulation, ECG characteristic signals are evaluated accurately, this algorithm laid good foundation for follow-up study of ECG.

关 键 词：心电信号 小波变换 多孔算法 Marr小波 QRS波群 

分 类 号：R318.0[医药卫生—生物医学工程] TP391[医药卫生—基础医学]