北京生物医学工程

心音信号分段规则的探讨

Discussion on the segmentation rules for heart sound signals

作者：吴玉春杨亚萍邓北川

单位：                      西安航空学院（西安710077）        

关键词：心音信号；分段；自动识别

分类号：

出版年·卷·期（页码）：2014·33·5（524-526）

摘要：

目的如何有效提取心音信号的有效成分(第一心音S1、第二心音S2)是分析心音信号的关键。为提取心音信号的有效成分，必须明确心音信号的分段规则。方法首先对目前心音研究领域中常用的两类心音分段方法进行分析和比较。根据现有文献资料，结合作者对实际采集的829例心音实例的研究，提取心音的各时域特征并进行统计分析，最后对心音信号分段规则进行了探讨。结果心音信号的S1、S2及收缩期、舒张期等时域特征呈现一定规律性。结论可按照上述时域特征对心音信号进行自动分段，并借助心音分段规则进行进一步识别和分析。

Objective The key point of analyzing heart sound signal is how to extract the first heart sound and the second heart sound (S1, S2). In order to extract the active ingredients of heart sound signals, segmentation rules must be clear. Methods Firstly, Two methods of heart sound segmentation are analyzed and compared. By studying some document literature and 829 heart sound samples collected by the authors, we extract heart sounds’ time domain features and do some statistical analysis. Finally, we discuss the segmentation rules for heart sound signals. Results S1, S2, systolic, diastolic and other time-domain characteristics of heart sound signals show certain regularity. Conclusions Time domain characteristics can be used to automatically segment heart sound signals, and the segmentation rules for heart sound are helpful to further identify and analyze the signals.

参考文献：

［1］吴延军,徐泾平,赵艳.心音的产生与传导机制［J］.生物医学工程学杂志,1996,13（3）:280-288.
Wu Yanjun, Xu Jingping, Zhao Yan, et al. The mechanism of generation and transmission of heart sound ［J］. Journal of Bio-medical Engineering, 1996, 13(3): 280-288.
［2］Boutemedjet S, Bouguila N, Ziou D. A hybrid feature extraction selection approach for high-dimensional non-gaussian data clustering［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2009, 31(8): 1429-1443.
［3］Groch MW, Domnanovich JR, Erwin WD. A new heart sounds gating devices for medical imaging［J］. IEEE Trans on Biomedical Engineering, 1992, 39(5):307-310.
［4］Gerbarg DS. Analysis of phonocardiogram by a digital computer［J］. Circulation Research, 1962,11:569-576.
［5］Quan Haiyan, Wang Weilian. Extraction of the first and the second heart sounds based on multi-resolution analysis of wavelet transform［J］. Beijing Biomedical Engineering, 2004:23(1):64-66.
［6］林勇,许晓飞.基于经验模式分解的心音自动分段算法［J］.中国生物医学工程学报,2008,27(4):485-489.
LinYong, Xu Xiaofei. Segmentation algorithm of heart sounds based on empirical mode decomposition［J］. Chinese Journal of Biomedical Engineering, 2008, 27(4):485-489.
［7］周静,杨永明,何为.心音信号的分析及其特征提取方法的研究［J］.中国生物医学工程学报,2005,24(6):685-689.
Zhou Jing, Yang Yongming, He Wei. A new algorithm of heart sound feature extraction［J］. Chinese Journal of Biomedical Engineering, 2005, 24(6):685-689.
［8］Ricke AD, Povinelli RJ, Johnson MT. Automatic Segmentation of Heart Sound Signals Using Hidden Markov Models［J］. Computers in Cardiology,2005,32:953-956.
［9］陈萌辉,叶大田,陈江天.基于信号包络及短时过零率的心音分段算法［J］.北京生物医学工程, 2007, 26(1):48-51.
Chen Menghui, Ye Datian, Chen Jiangtian. Heart sound segmentation algorithm based on shannon energy and short-time zero-crossing rate［J］. Beijing Biomedical Engineering, 2007, 26(1):48-51.
［10］Liang H, Lukkarinen S, Hartimo I. Heart sound segmentation algorithm based on heart sound envelogram［J］.Computers in Cardiology,1997,24:105-108.
［11］Samit Ari, Goutam Saha. On a robust algoritm for heart sound segmentation［J］. Journal of Mechanics in Medicine and Biology,2007,7(2):129-150.
［12］Hebden JE, Torry JN. Neural network and conventional classifiers to distinguish Between first and second heart sounds［J］.IEEE Colloquium on Artificial Intelligence Methods for Biomedical Data Processing,1996,26(4):1-6.
［13］屠志海.心音信号的分析方法研究［D］.上海:华东师范大学, 2011.
Tu Zhihai. Research on analysis for heart sounds［D］. Shanghai: East China Normal University, 2011.
［14］刘娟. 噪声环境下心音身份识别算法研究［D］. 杭州:杭州电子科技大学,2010.
Liu Juan. Study on heart sound identification algorithm in noise background［D］. Hangzhou: Hangzhou Dianzi University,2010.
［15］吴玉春.心音信号识别研究方法进展［J］.北京生物医学工程,2012，31（3）：325-330.
Wu Yuchun.Progress on heart sound recognition Methods［J］. Beijing Biomedical Engineering, 2012，31（3）：325-330.

服务与反馈：

【文章下载】【加入收藏】

提示：您还未登录，请登录！点此登录