基于对数域连续小波变换电路的心电图QRS波检测  

Detecting QRS of ECG Based on Log-domain Continuous Wavelet Transform Circuits

作　　者：李宏民[1] 何怡刚[1] 唐圣学[1] 刘美容[1] 张颖[1] 

机构地区：[1]湖南大学电气与信息工程学院

出　　处：《中国生物医学工程学报》2008年第1期8-12,共5页Chinese Journal of Biomedical Engineering

基　　金：国家自然科学基金资助项目(50677014);国家863计划(2006AA04A104);湖南省科技计划项目(06JJ2024);教育部新世纪优秀人才支持计划(NCET-04-0767);高校博士点基金(20060532002)

摘　　要：为满足植入式心脏起搏器之类的医疗设备低功耗、实时处理等应用要求的需要,提出了基于低电压、低功耗对数域连续小波变换电路的心电图QRS波检测方法。为便于用模拟VLSI实现小波变换,用混合粒子群算法构造了类高斯一阶导数小波。以平衡式对数域积分器为积木块,设计了用于QRS波检测的连续小波变换电路,该电路由冲激响应为类高斯一阶导数小波函数的反褶及其伸缩的滤波器组构成。由该电路实现心电信号的小波变换,进行QRS波检测。仿真结果表明了该方法的可行性。To meet the requirement of low-power, low-voltage and real-time in the medical devices such as implantable pacemakers, a novel method for detecting QRS of ECG based on log-domain analog VLSI implementing continuous wavelet transform was presented. A new wavelet so called similar first derivative of the Gaussian wavelet was constructed using a hybrid particle swarm algorithm. Then, continuous wavelet transform circuits were designed. The circuits were composed of analog filters, whose impulse response was the constructed wavelet. The filter design was based on balanced log-domain integrators. By implementing wavelet transform of ECG in this circuits, QRS complex was distinguished in a low power and real time environment. The validity of the proposed method was confirmed by simulation results.

关 键 词：QRS波 小波变换 模拟VLSI 对数域滤波器 优化算法 

分 类 号：R318.04[医药卫生—生物医学工程]