双声道心音能量熵比的提取与识别研究<br />-《计算机技术与发展》

文章信息/Info

Title:: Research on Feature Extraction and Identification with Specific Entropy of Dual Track Heart Sound

作者:: 孙科学1 2 ; 魏敏1 ; 宋江珉1 ; 李子炎1 ; 成雨含3; 1. 南京邮电大学电子科学与工程学院,江苏南京 210023;
2. 江苏省射频集成与微组装工程实验室,江苏南京 210023;
3. 南京邮电大学信息化建设与管理办公室,江苏南京 210023

Author(s):: SUN Ke-xue 1 2 ; WEI Min 1 ; SONG Jiang-min 1 ; LI Zi-yan 1 ; CHENG Yu-han 3; 1. School of Electronic Science and Engineering,Nanjing University of Posts and Telecommunications,Nanjing 210023,China;
2. Jiangsu Province Engineering Lab of RF Integration &Micropackage,Nanjing 210023,China;
3. Office of Informatization and Management,Nanjing University of Posts and Telecommunications, Nanjing 210023,China

Keywords:: dual track heart sound; energy entropy ratio; similarity distance; classification recognition

摘要:: 研究了双声道心音的能量熵比提取与分类识别的基本原理和实现方法。分析了心音信号特性及提取双声道心音
特征值的可行性和必要性。为了减少特征值的计算量,又能使特征值体现出时域和频域的不同特点,提出了双声道心音
能量熵比的概念并详细地阐述了其计算过程。分别取单声道心音时域能量熵比和频域能量熵比,双声道心音的能量熵比
为两个单声道心音能量熵比的均值。将双声道的能量熵比作为特征值识别双声道心音。然后通过相似距离法针对双声
道心音能量熵比进行双声道心音的分类识别。实验结果表明,双声道心音能量熵比能全面地反映双声道心音的时频域特
点,不仅有利于心音的分类识别,而且降低了特征值的维度,使得计算量减小,提高了识别效率和实用性。

Abstract:: The fundamental principle and implementation method of extracting energy entropy ratio and classification recognition from the
dual track heart sound are researched. The feasibility and necessity of characteristics of heart sound and extracting dual track heart sound
features have been analyzed. In order to reduce the calculation of eigenvalue and show its characteristics in time and frequency domain,
the concept of energy entropy ratio for dual track heart sound is presented and its computational process is discussed in detail. After calcu-
lating the energy entropy ratio of the mono heart sounds in time and frequency domain,the dual track heart sound energy entropy ratio is
the average of them. Then the method of similarity distance is applied to classify and recognize the dual track heart sound. The experimen-
tal results show that the characteristics in time-frequency domain of dual track heart sound can be reflected comprehensively by the ener-
gy entropy ratio,which is not only beneficial to the recognition of heart sound,reducing the eigenvalue dimensions,but also make it possi-
ble to reduce amount of calculation and improve the recognition efficiency and practicality.