基于数学形态学的心音包络提取与识别方法研究.pdfVIP

第 23 卷 　第 3 期 北京生物医学工程 Vol23 　No3 2004 年 　9 月 Beijing Biomedical Engineering Sep . 　2004 基于数学形态学的心音包络提取与识别方法研究 姚晓帅 　郭兴明　陈剑 　肖守中 　　摘 　要 　心音包络比原始心音可以更好地显示心音的特征 , 是进行心音独立识别的基础 。本文把数学形态 学应用于心音包络的提取和识别的研究 。首先利用形态学滤波和全波整流对原始心音进行预处理 ; 然后利用形 态学闭运算提取心音包络 ; 最后应用形态学开运算来消除噪声包络 。在数学形态学提取的心音包络基础上 , 对 50 例心音样本进行了第一心音 、第二心音识别 , 全部 20 例正常心音的第一心音和第二心音被正确识别 , 27 例包 含心杂音的异常心音的第一心音 、第二心音也被正确识别 。为进一步的心音分析及心音诊断奠定了基础 。 关键词 　心音 　数学形态学 　包络提取 　数学形态学滤波 中图分类号 　R31804 文献标识码 A 文章编号 (2004) Envelope Extraction and Recognition of Heart Sounds Based on Mathematical Morphology 　 YA O X iaoshuai , GUO X ingming , CHEN J ian , XIA O S houzhong 　College of B ioengineering , Key L aboratory f or B iomechanics Tissue Engineering ( Chongqing University ) , Ministry of Education , Chongqing 　400044 In this paper , a new method was put forward for envelope extraction and recognition of heart sounds based on mathematical 【Abstract 】　 morphology . Firstly , the original heart sounds signal was preprocessed by mathematical morphology filter and fullwave rectification ; Secondly , the envelope of heart sounds was extracted by closing operation of mathematical morphology ; Finally , the envelope of noise included in the heart sounds was eliminated by opening operation of mathematical morphology . And based on the envelope extracted the first and second heart sounds of 50 samples of heart sounds were recognized . The result showed that all of 20 samples of normal heart sounds were recognized and 27 samples of heart sounds , including murmurs , were recognized