基于匹配追踪算法的信号复杂度在心音分段定位中的应用

Application of Matching Pursuit Based on Signal's Relative Structural Complexity in the Segmentation of Heart Sound

分析心音信号的关键是如何有效地提取心音信号的有效成分(第一心音S1、第二心音S2)。文中结合心音信号的自身特点,把心音信号看成含有冲击成分的冲击信号,将匹配追踪算法引入到心音信号分析中,基于该算法提出利用相关结构复杂度对心音信号进行特征参数的提取方法。首先将心音分段,然后计算每段信号的分解次数,将分解次数定义为信号的相关结构复杂度。根据复杂度曲线可以有效地定位出S1、S2,还可以提取心音信号的一些典型特征参数。

The key point of analyzing heart sound signal is how to extract the first heart sound and the sec- ond heart sound （S1, S2）. According to the characteristics of heart sound in this paper, the signal is con- sidered as shock signal and the matching pursuit algorithm is used to analyze the heart sound signal. We use relative structural complexity to extract the features of heart sound based on the matching pursuit algo- rithm. Firstly, we segment the heart sound, then calculate the relative structural complexity of every sec- tion. The characteristics of heart sound can be extracted effectively by using the proposed method.