摘要
针对传统小波变换的去噪算法运算复杂,难以用于心电信号的实时处理,而常见心电信号滤波算法在实际应用中去噪效果不理想,为正确识别心电信号,提高实时性和去噪效果,采用提升方案来构造小波,提高了小波分解的速度,减少算法对内存的需求,并结合阈值滤波算法对小波系数进行处理,实现信号与噪声的分离。为了验证算法有效性,对MIT-BIH数据库中数据进行了仿真实验,结果表明方法处理后信号失真较小,信号中叠加的工频干扰和肌电干扰基本被消除,相对于基于传统小波变换算法处理速度有了很大的提升。
De-nosing methods based on wavelet transform cannot be applied to real-time ECG signals processing due to computation complexity,and some traditional algorithms cannot provide desired effect in actual use.In this paper lifting scheme is used to speed up wavelet decomposition,to reduce memory demand,and combining with threshold de-nosing method to deal with noised ECG signals.To validate the actual effect of the method,signals in MIT-BIH ECG signal database are analyzed.Results show that the de-noised signal has little distortion,and the added electrode motion artifact and muscle artifact are successfully removed,the computing speed promotes significantly comparing with algorithm using wavelet transform.
出处
《计算机仿真》
CSCD
北大核心
2010年第10期226-229,共4页
Computer Simulation
基金
国家自然科学基金(60874005)
中澳国际合作项目特别资金(071107037)
关键词
提升方案
去噪
心电图
工频干扰
肌电干扰
Lifting scheme
De-Noising
ECG
Electrode motion artifact
Muscle artifact
