心音信号识别研究方法进展

有效提取心音信号中的第1、2心音（S1和S2）是心音信号研究的关键点.本文介绍了心音信号识别的研究现状,通过比较谱分析、小波变换、神经网络以及数学形态学方法等几种心音研究方法,讨论了心音信号识别的研究重点和发展方向,为进一步的心音研究奠定基础.     

心音信号识别的意义及其方法的研究

本文概述了心音信号识别的意义,并对心音自动识别技术的发展进行了介绍,最后总结了今后工作可能的研究方向。     

基于STMHT算法的心音分割研究

第一心音（S1）和第二心音（S2）的定位和提取是利用心音分析诊断心脏病时的首要任务。鉴于此,本研究提出一种基于STMHT的心音分割法,分别提取S1和S2。本研究分为以下3个阶段:第一阶段,采用小波分解对心音信号进行预处理,保留心音信号的有效成分（21.5~689.0 Hz）;第二阶段,用Viola积分波形法提取心音包络;最后,基于STMHT算法自动定位和提取S1和S2。对30例心音信号的提取结果进行评价,结果表明,S1和S2提取的准确率高达97.37%,优于其它已实现的有效方法。     

心音信号分段规则的探讨

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。     

基于优化后向传播神经网络的基础心音分类

针对后向传播（BP）神经网络高度依赖初始权值、收敛慢且易陷入局部极值,标准人工蜂群算法开发能力弱、局部 搜索能力差等问题,提出一种基于改进人工蜂群算法优化BP神经网络的分类方法。引入自适应和全局最优策略改进人 工蜂群算法中跟随蜂蜜源全局搜索、概率选择算法,使用当前迭代的最优解来提高其开发能力。此外,利用混沌系统产生 初始种群,以增强人工蜂群算法全局收敛性。最后,将本文算法应用到基础心音分类。结果表明本文算法较经典分类算 法分类准确率有较大的提升。梅尔频率倒谱特征参数下,本文算法的分类准确率达到94%以上。     

基于周期提取的心音分段方法

目的心音分段是心音分析的基础,传统方法是利用心音基本成分进行识别,而病变的心音信号中含有的杂音使识别受到干扰,易产生误分段。本文提出了基于周期提取的信号分段方法,可以避免对心音基本成分的识别。方法以虚拟仪器Lab VIEW为开发平台,首先利用小波变换对原始心音进行去噪预处理,然后利用快速Hilbert变换提取心音包络,再利用其自相关分析函数求出心动周期,进而从原始心音信号中提取整周期的信号,避免对心音基本成分的识别。结果对30例心音信号做实验验证,得到的心动周期长度能够直观显示,正确率达98%以上。结论作为一种无需识别心音基本成分的分段方法,此方法为后续的特征提取等研究打下了坚实基础。     

正常与异常心音信号比较研究

目的：获取心音信号中对心脏疾病诊断有意义的信息。材料与方法：利用自回归模型方法对１６９个受检者（包括３组正常人和４组心脏病患者）的第１－４心音和收缩人期杂音、舒张杂音信号进行了谱分析。并各提取了５项对疾病诊断有意义的特征参数（频域３项，时域２项）进行正常与正常心音信号的比较研究。结果：结果表明时，频域许多参数均有显著性差异。结论：研究结果可为心音研究的进一步深入和临床心脏病的辅助诊断提供基础数据和     

Reduction of heart sound interference from lung sound signals using empirical mode decomposition technique

During the recording time of lung sound (LS) signals from the chest wall of a subject, there is always heart sound (HS) signal interfering with it. This obscures the features of lung sound signals and creates confusion on pathological states, if any, of the lungs. A novel method based on empirical mode decomposition (EMD) technique is proposed in this paper for reducing the undesired heart sound interference from the desired lung sound signals. In this, the mixed signal is split into several components. Some of these components contain larger proportions of interfering signals like heart sound, environmental noise etc. and are filtered out. Experiments have been conducted on simulated and real-time recorded mixed signals of heart sound and lung sound. The proposed method is found to be superior in terms of time domain, frequency domain, and time-frequency domain representations and also in listening test performed by pulmonologist.     

Reduction of heart sound interference from lung sound signals using empirical mode decomposition technique

During the recording time of lung sound (LS) signals from the chest wall of a subject, there is always heart sound (HS) signal interfering with it. This obscures the features of lung sound signals and creates confusion on pathological states, if any, of the lungs. A novel method based on empirical mode decomposition (EMD) technique is proposed in this paper for reducing the undesired heart sound interference from the desired lung sound signals. In this, the mixed signal is split into several components. Some of these components contain larger proportions of interfering signals like heart sound, environmental noise etc. and are filtered out. Experiments have been conducted on simulated and real-time recorded mixed signals of heart sound and lung sound. The proposed method is found to be superior in terms of time domain, frequency domain, and time–frequency domain representations and also in listening test performed by pulmonologist.     

心音信号的时频分析

对比研究了几种时频分析方法(WVD、CWD、CKD)及其在心音信号时频分析方面的特征和差异，并分别对正常和病态的一个心动周期的心音时频特征进行了研究，结果显示时频分析方法对心音这种非平稳信号，有较高的时频分辨率，这对于揭示心音产生的生理机制有着积极的作用，在理论研究和临床诊断中有一定的实用价植。     

A robust heart sound segmentation algorithm for commonly occurring heart valve diseases

The first step towards detection of valvular heart diseases from heart sound signal (phonocardiogram) is segmentation. A segmentation algorithm provides the location of the first and second heart sounds which in turn helps to locate and analyse the murmur. Established phonocardiogram based segmentation methods use an electrocardiographic (ECG) signal as a continuous auxiliary input in a complex instrumentation setup. This paper proposes an automatic segmentation method that does not require any such auxiliary signal. Compared to other approaches without auxiliary signal, this work extensively utilizes biomedical domain features for reduction of time and computational complexities and is more accurate. The performance of the algorithm is evaluated for nine commonly occurring pathological cases and normal heart sound for various sampling frequencies, recording environments and age group of subjects. The proposed algorithm yields an overall accuracy of 97.47% and is compared with two competing techniques. In addition, the robustness of the algorithm is shown against additive white Gaussian noise contamination at various SNR levels.     

A robust heart sound segmentation algorithm for commonly occurring heart valve diseases

The first step towards detection of valvular heart diseases from heart sound signal (phonocardiogram) is segmentation. A segmentation algorithm provides the location of the first and second heart sounds which in turn helps to locate and analyse the murmur. Established phonocardiogram based segmentation methods use an electrocardiographic (ECG) signal as a continuous auxiliary input in a complex instrumentation setup. This paper proposes an automatic segmentation method that does not require any such auxiliary signal. Compared to other approaches without auxiliary signal, this work extensively utilizes biomedical domain features for reduction of time and computational complexities and is more accurate. The performance of the algorithm is evaluated for nine commonly occurring pathological cases and normal heart sound for various sampling frequencies, recording environments and age group of subjects. The proposed algorithm yields an overall accuracy of 97.47% and is compared with two competing techniques. In addition, the robustness of the algorithm is shown against additive white Gaussian noise contamination at various SNR levels.     

一种基于短时傅立叶变换的心音分裂时间计算方法

目的应用现代信号处理的方法定量计算心音分裂的时间间隔，为某些心脏早期器质性病变的诊断提供数据依据。方法在频率分辨率较高的情况下，利用短时傅立叶变换（STFT）声谱图和香农能量，提取出第1心音（S1）的主要成分二尖瓣关闭音（M1）、三尖瓣关闭音（T1）及第2心音（S2）的主要成分主动脉瓣关闭音（A2）、肺动脉瓣关闭音（P2）。然后，在时间分辨率较高的情况下，通过瞬时能量密度包络图，计算出心音分裂的时间间隔。结果对南开大学医学院提供的心音数据的仿真结果表明该方法能够较精确地计算出房间隔缺损（ASD）、右束支传导阻滞（RBBB）及其他常见心音分裂类型的分裂时间。结论笔者提出的计算心音分裂时间间隔的方法比已有的方法更简单快捷，其结果能够为某些心脏早期器质性病变的诊断提供定量依据。