Detection of Mitral Valve Diseases by Bicoherence Analysis of Heart Sound Signals

The analysis of 10 normal and 51 mitral valve pathology making a total of 61 heart sound signals that were obtained with Littmann 4100 Digital Stethoscope were conducted in this study.Following the recorded sound signals were denoised by using wavelet filters,the signals were applied bicoherence analysis that is an high order spectral analysis method.It has been demonstrated that varieties of mitral valve pathology could be determined by three-dimensional surfaces of bicoherence and maximum bicoherence values.

Research on Heart Sound Signal Denoising Algorithm Based on Variational Mode Decomposition and Wavelet Threshold

Heart sound signals are easy to introduce noise during the acquisition process, and traditional denoising algorithms always remove the characteristic information of the heart sound while removing the noise. The denoising effects in turn affect the subsequent diagnosis results. So an improved algorithm based on variational mode decomposition (VMD) and wavelet threshold method is proposed. First, the number of decomposition modes K of the VMD is selected by analyzing the average instantaneous frequency curve of the different decomposition values, and the noisy heart sound is decomposed into K modes by the VMD algorithm. Then, the modes that need to be retained are decided by the energy curve of each mode. Finally, wavelet threshold denoising method is performed on the retained modes. Experiment simulation results show that under different signal-to-noise ratio conditions, the proposed method can improve heart sounds’ ratio of signal to noise and reduce the root mean square error. Compared with traditional algorithms, it has good noise suppression capabilities under different noise levels.

Development of laser-based heart sound detection system

In this paper, we demonstrate the prototype of a new stethoscope using laser technology to make the heart-beat signal “visible”. This heartbeat detection technique could overcome the limitation of the acoustic stethoscope brought by the poor ability of human ear to hear low frequency heart sounds. This is important, as valuable information from sub-audio sounds is present at frequencies below the range of human hearing. Moreover, the diagnostic accuracy of the acoustic stethoscope is also very sensitive to noise from immediate environment. In the prototype of laser-based stethoscope, the heartbeat signal is correlated to the optical spot of a laser beam reflected from a thin mirror attached to the patient’s chest skin. The motion of the mirror with the chest skin is generated by the heart sounds. A linear optical sensor is applied to detect and record the motion of the optical spot, from which the heartbeat signal in time-domain is extracted. The heartbeat signal is then transformed to frequency domain through digital signal processing. Both time-domain and frequency-domain signals are analyzed in order to classify different type of heart murmurs. In the prototype of the laser-based stethoscope, we use a heart-sound box to simulate the chest of a human being. The heart-sounds collected from real patients are applied to activate the vibration of the heart-sound box. We also analyze different heart murmur patterns based on the time-domain and frequency-domain heartbeat signals acquired from the optical system.

Machine Learning-Based Intelligent Auscultation Techniques in CongenitalHeart Disease: Application and Development

Congenital heart disease(CHD),the most prevalent congenital ailment,has seen advancements in the“dual indi-cator”screening program.This facilitates the early-stage diagnosis and treatment of children with CHD,subse-quently enhancing their survival rates.While cardiac auscultation offers an objective reflection of cardiac abnormalities and function,its evaluation is significantly influenced by personal experience and external factors,rendering it susceptible to misdiagnosis and omission.In recent years,continuous progress in artificial intelli-gence(AI)has enabled the digital acquisition,storage,and analysis of heart sound signals,paving the way for intelligent CHD auscultation-assisted diagnostic technology.Although there has been a surge in studies based on machine learning(ML)within CHD auscultation and diagnostic technology,most remain in the algorithmic research phase,relying on the implementation of specific datasets that still await verification in the clinical envir-onment.This paper provides an overview of the current stage of AI-assisted cardiac sounds(CS)auscultation technology,outlining the applications and limitations of AI auscultation technology in the CHD domain.The aim is to foster further development and refinement of AI auscultation technology for enhanced applications in CHD.

The analysis of heart sounds by using bispectral technique

An advanced signal processing technique, higher-order spectra, is proposed to in vestigate the nonlinear coupling phenomena of the heart sounds. To extract more higher-order information of the heart sounds, a non-Gaussian AR model is selected for parametric bispectral estimation in analyzing several kinds of heart sounds. The non-Gaussian AR model of the sound signals is llsed to detect quadratic nonlinear interactions and to classify two patterns of heart sounds in terms of the parametric bispectral estimate. The bispectral cross-correlation is employed to the order determination of the model. Several real heart sound data are imple mented to show that the quadratic nonlinearity exist in both normal and clinical heart sounds.It was found that bispectral techniques are effective and useful tools in analyzing heart sounds and other acoustical signals

心音频谱图在二尖瓣反流患者左心室功能监测中的作用

目的探求使用可穿戴设备获取心音频谱图,观察不同程度二尖瓣反流(MR)患者心音频谱特征在左心室功能监测中的作用。方法入选2020年6月至2021年5月因心血管疾病入住上海交通大学医学院附属瑞金医院的102例MR患者,根据心脏超声MR反流束面积/左心房面积比值超过20%将其分为轻度MR组72例和中度及以上MR组30例。基于心音频谱图对心音持续时间、最大声强、最大心音振幅进行分析,并与心脏超声测定的左心室射血分数(LVEF)进行相关性分析。结果心音频谱图第一心音持续时间与收缩期持续时间比值(S11-S12/S11-S21)与LVEF呈线性负相关(r=-0.535,P<0.01),S11-S12/S11-S21用于判定LVEF降低(LVEF<50%)曲线下面积为0.840(95%CI=0.756~0.924,P<0.01),截断值是55.50%,敏感度78.6%,特异度83.3%。结论心音频谱图提示,MR患者心脏收缩期第一心音持续时间占比与LVEF呈负相关性,在二尖瓣病患的心功能居家监测中存在潜在的应用价值。

基于活动性检测动态估计噪声的心音降噪算法

针对基于小波分解和最优改进对数幅度谱估计的心音降噪算法存在噪声残留和心音失真的问题,提出一种基于心音活动性检测(HSAD)动态估计噪声的心音降噪算法。通过设计的HSAD判断当前心音帧是否为基础心音帧(FHS),根据判断结果分别采用改进最小值控制递归平均(IMCRA)算法和递归平滑算法对噪声功率进行动态估计与更新,采用非因果先验信噪比,实现心音信号的降噪。实验结果表明,提出算法能更好在提升降噪性能的同时,降低FHS的失真。

基于时频融合特征的肺动脉高压心音分类模型

先心病相关肺动脉高压是一种严重的心血管疾病,致死率高,对其进行早期筛查与识别对于治愈尤为重要。目前临床是通过右心导管术确诊,此为有创检查,不便于在大规模筛查中采用,研究一种无创便捷的识别方法迫在眉睫。文中建立了一种时频融合的心音分类模型。首先对心音信号进行预处理,然后使用融合滤波器组对信号进行转换并求取动态时频特征,最后将得到的融合特征参数输入表格式先验数据拟合网络(TabPFN)中进行分类识别。实验结果表明,该算法在正常、CHD-PAH和CHD中的平均准确率、精确率、灵敏度、特异度和F1分别为92.21%,92.15%,92.15%,96.11%,92.14%。对于先心病相关肺动脉高压的早期筛查与识别具有重要意义。

心音心电检测仪电磁兼容问题的测试分析及整改方法

现代医疗器械更偏向于高精确度、高敏感度和小型化,其所处的电磁环境越来越复杂。外界电磁干扰对医疗设备会产生影响,严重时甚至影响医生的诊断。本文结合心音心电检测仪,基于医疗器械测试标准对传导发射测试、辐射发射测试和静电放电这3个测试项目进行了研究,提出了壳体整改,增加滤波电路、磁环、电荷释放电路等整改方法。对其他类型相关医疗器械的电磁兼容问题分析和整改,有一定的参考借鉴作用。

MEMS心音传感器及检测电路优化设计

对基于微电子机械系统(MEMS)技术的心音传感器声敏结构进行了优化且设计了其检测电路。首先,针对心音信号的特点,设计了二次集成的扁平状仿生纤毛结构,对该结构进行仿真,确定了纤毛的尺寸参数和梁上最大应力1.2×10^(5) N/m^(2),对纤毛进行特征频率仿真,在硅油域中结果为711 Hz;其次由扁平状纤毛结构X轴接收噪声时梁上的应力仿真结果可知该结构具有抗干扰能力;最后设计了后端的放大电路和滤波电路并对传感器封装后进行测试。测试结果表明,该结构的信噪比达到了27 dB,较传统的圆柱形纤毛提高了35%,且其抗干扰能力也优于传统的圆柱形纤毛。优化过后的MEMS心音传感器具有抗干扰、低噪声、低成本、采集的信号不失真等优势,可为临床心音信号的采集提供关键核心部件。

改进变值逻辑与线性预测在心音分类中的应用

心音对于评价心脏健康状况具有重要作用.文章介绍了一种新的基于变值逻辑与线性预测倒谱系数融合特征的先心病分类算法,有助于提取心音中的深度病理特征.算法首先对心音进行降噪、包络提取;然后进行变值逻辑运算、标记并转换为可分析的测度数据,并计算信号的线性预测倒谱系数进行特征融合;最后使用随机森林,XGBOOST和LIGHTGBM机器学习分类器进行先心病二分类.研究所用心音样本共4000例,测试结果对正常和异常心音分类的平均准确率为0.9138.算法无需对心音进行心动周期分割,大大简化了分析流程,可望用于先心病的筛查.

面向可穿戴式的基于LSTM神经网络的智能心音异常诊断芯片

心血管疾病是造成全球死亡人数最多的疾病之一,因此对心血管疾病的预防与提前诊断至关重要。人工听诊技术与计算机心音诊断技术无法满足对心音长时间听诊的需求,因而可穿戴式听诊设备越来越受到关注,但是其具有高精度与低功耗的要求。该文设计了低功耗的面向可穿戴式的基于长短期记忆网络(Long Short-Term Memory, LSTM)的智能心音异常诊断芯片,提出了包括预处理、特征提取以及异常诊断的心音异常诊断系统,并搭建了基于听诊器的心音采集FPGA系统,采用了数据增强的方法解决数据集的不平衡问题。基于预训练模型设计了智能心音异常诊断芯片,在SMIC180 nm工艺下完成了版图设计和MPW流片。后仿真结果表明,智能心音异常诊断芯片的诊断准确率为98.6%,功耗为762μW,面积为3.06 mm×2.45 mm,满足可穿戴式智能心音异常诊断设备的高性能与低功耗的需求。

心电心音同步检测的临床应用研究进展

心音是由心肌收缩、心脏瓣膜关闭和血液撞击大动脉壁、心室壁等引起的振动所产生的机械波,心血管血流动力学改变及心脏结构异常往往伴随着心音的变化。心脏听诊是获取心音最直接的方式,但结果受医生听诊经验的影响较大。心音图(phonocardiogram,PCG)是将心音的振动转变为时间序列振动的波形图。心电心音同步检测(acoustic cardiography,ACG)是一种基于心电心音信号同步采集并分析的新型计算机辅助诊断技术,近年来,在临床上的应用特别是在心血管疾病的筛查、诊断及预后中取得了很大的进展。本文对ACG在临床上的应用研究进展作一综述。

心肺音分离方法研究进展

听诊是诊断心血管和呼吸系统疾病最有效的方法。为了达到准确诊断的目的,设备必须能够识别各种临床情况下的心肺音。然而,记录的胸腔声音通常为心肺音混合信号。因此,将心肺音混合信号分离对于医生听诊至关重要。本文介绍了心音信号和肺音信号的频率范围和信号特征,综述了目前心肺音分离方法的研究进展,阐述了现有各种心肺音分离方法的优缺点,指出了选取一种合适的心肺音分离算法来分离心肺音对辅助医疗的重要意义。

基于迁移学习的心音心脏疾病检测算法

心脏疾病一直是全球范围内重要的健康挑战,早期的诊断和检测至关重要。研究旨在开发一种基于迁移学习的心音心脏疾病检测算法,以提高心脏疾病的早期诊断准确性。训练时使用迁移学习增强了模型的稳定性和泛化能力。另外,提出一种基于HTS-AT模型的改进的心音识别模型HTS-AT V2,对心音信号进行特征提取和分类。结果表明,改进的算法在心音检测方面取得了显著成功,在检测效果有提升的同时,加快了推理速度、减少了模型大小。

遗传算法与修正的自适应矩估计优化循环神经网络的心音分类方法

针对传统的循环神经网络(RNN)在识别分类心音信号方面具有梯度爆炸、梯度消失和短期记忆的问题,该文提出了无需心音分段的结合遗传算法(GA)与修正的自适应矩估计(RAdam)优化RNN的心音分类模型。该模型的优势是将GA和RAdam优化器以串联的方式融合到RNN中,以达到改进RNN的作用。首先,利用GA的选择、变异和遗传操作,优化RNN的输入层节点数,获取心音特征向量的最优个体的初始解。其次,根据最优个体中的权重、偏置矩阵,赋予模型初始权值和阈值,获得初始权重最优解,整个模型共享参数。最后,联合改进的学习率自适应优化算法,优化RNN模型。结果表明,结合经典的梅尔(Mel)倒频谱系数方法提取心音信号的特征向量,心音信号分类准确率达到90.29%,相比于未优化的RNN模型,准确率提高了17.79%。

心音特征智能分析在心血管功能评估和疾病诊断中的应用

数字化心音与人工智能技术的结合可实现对心音进行精确地连续、定量分析和分类识别,使心音特征提取与融合分析在心脏血流动力学监测、心力衰竭分型诊断、先天性或风湿性心脏病分类、冠状动脉疾病检测等领域成为研究热点。本文综述心音特征智能分析在心血管功能评估和疾病诊断中的应用。

基于小波包重构信号能量分布特征的心音分类识别

目的:为了有效识别心脏疾病心音的病理特征信息进行心脏疾病早期筛查,提出一种基于小波包系数重构信号能量序列的分布特征提取算法。方法:应用小波包分解算法对原始心音信号进行10层成分分解,获得各层小波包系数后对每一个系数进行重构,计算重构信号的能量并按原序排列构成能量序列。分析各层重构信号的能量序列的分布特征,并把分布特征作为分类特征。应用支持向量机、K近邻和决策树对正常心音和各类心脏疾病心音信号进行分类识别。结果:应用重构信号能量序列的分布特征结合决策树分类器,对公开数据集的5种心音分类识别准确率可达93.6%;对临床采集的正常心音和肥厚性心肌病心音数据分类准确率最高达95.6%。结论:本文算法能提取异常心音信号的有效病理信息,为临床心脏病听诊提供参考。

基于改进NMF与相位补偿的胎心音降噪算法

在孕妇体表,通过声音传感器对胎儿心音信号进行采集,结果通常包含多种噪声,常用的滤波方法无法取得满意的降噪效果。为了能够更好地滤除与胎儿心音信号同频的噪声信号,提出改进的非负矩阵分解算法(NMF),对短时幅度谱利用奇异值分解得到用于NMF的特征数和初始化矩阵,对NMF增加L2,1稀疏限制,使分解得到的胎心音特征呈现更多细节;对短时相位谱采用了增加稀疏限制的相位补偿算法(PSC)。实验结果表明:与常用降噪方法相比,该方法的信噪比至少提升0.52 dB,能更好保留胎儿心音信号特征。