心音信号周期增量序列的多尺度化研究

Research on Multiscale Analysis of Heart Sound Cycle Increment Series

心音信号是一种复杂的生理信号,对心音信号产生机理的研究能够为心音听诊提供理论依据。文中从心血管循环系统的生理结构出发,讨论了心音的产生机理;然后,提出心音周期增量序列的多尺度化基本尺度熵和相关的评价指标;最后,利用本方法对健康人群和充血性心力衰竭患者进行分类分析。仿真实验表明,对心音周期增量序列进行多尺度化基本尺度熵分析,可以准确地区分健康人和心力衰竭人群。基于多尺度化基本尺度熵提出的诊断参数可以作为心衰早期诊断的一种依据。通过肩带式心音传感器首次实现了不用手持、长时间的心音采集,同时通过对长时间心音信号变化规律的研究,对于充分利用心音,挖掘听诊的潜力有着重要的意义。

Heart sound signal is a kind of complex physiological signals. The study on mechanism of heart sounds signal can provide theo-ry basis for heart sounds auscultation. In this paper,discuss the physical structure of cardiovascular circulation system and the generating mechanism of heart sounds. Then,propose multiscale base-scale entropy of heart sound cycle increment series and related evaluation in-dex. Finally,healthy people and patients with heart failure are diagnosed based on the proposed algorithm. Simulation experiments show that multiscale base-scale entropy of heart sound cycle increment series can accurately distinguish between healthy person and Congestive Heart Failure （ CHF） . The proposed diagnosis parameter based on multiscale base-scale entropy provides a new evidence for clinic of heart failure. Heart sounds can be collected for a long time without holding for the first time. The research on long heart sound signals is of great importance to make the best of heart sound and dig the great potential of heart sound.