小波消噪和ICA在语音信号分离中的应用

为了消除语音信号分离中仍存在的部分混叠声音,提出一种基于小波消噪和独立分量分析（ICA）结合的信号分离方法。该方法将小波变换和独立分量分析结合,利用小波变换的去噪作用,滤除原始语音信号中的噪声后作为ICA的输入信号,采用FastICA算法在小波域进行独立分量分析,对输入信号实施分离。实验结果表明,该方法大大调高了传统独立分量分析对语音信号的分离效果。     

ICA 在心音信号预处理中的应用研究

独立分量分析(ICA)是近年来涌现的用于盲信号分离的新技术，本文利用独立分量分析对心音信号进行了预处理：消除工频干扰。心音信号由自制的心音传感器获得。在分析了独立分量分析的基本原理的基础上，建立了基于互信息极小的目标函数，研究了目标函数优化的迭代算法，给出了利用此算法的ICA实现步骤。实验结果表明，利用独立分量分析有效地对心音信号进行预处理，能成功地从心音中分离出工频干扰信号。     

基于两级神经网络的心音分割

心音信号是分析诊断心脏疾病的重要信号,而心音分割是对其进行分析处理之前必不可少的一步。本文通过将心音分割任务分离为定位与识别两个子任务,提出一种两级卷积神经网络,由定位网络和判别网络两级构成,分别完成心音信号的识别与定位。首先将原始信号通过滑动窗口进行分帧,然后通过短时傅里叶变换得到其频谱,再通过梅尔滤波器得到其梅尔频谱系数（Mel frequency spectral coefficient, MFSC）特征,输入第1个定位网络对其是否为心音段进行判断,如果是的话,再输入判别神经网络,识别第一心音与第二心音,从而实现心音的分割。最后利用多帧结果投票,减小误判。同时,在卷积神经网络中引入空间注意力机制,实验结果表明,这种加入了注意力机制的两级神经网络模型在心音分割任务上比使用单个卷积神经网络分类模型的准确率更高,也使得模型更加简单,轻量化。     

基于HMM和WNN的心音信号身份识别研究*

将隐马尔可夫模型（HMM）与小波神经网络（WNN）相结合,提出了一种基于心音信号的身份识别方法。该方法首先利用HMM对心音信号进行时序建模,并计算出待识别心音信号的输出概率评分;再将此识别概率评分作为小波神经网络的输入,通过小波神经网络将HMM的识别概率值进行非线性映射,获取分类识别信息;最后根据混合模型的识别算法得出识别结果。实验采集80名志愿者的160段心音信号对所提出的方法进行验证,并与GMM模型的识别结果进行了对比,结果表明,所选方法能够有效提高系统的识别性能,达到了比较理想的识别效果。     

IMF复杂度特征在心音信号分类识别中的应用

为提高非线性、非平稳心音信号特征提取的准确性和分类识别的高效性,提出一种基于固有模态函数（Intrinsic Mode Function,IMF）复杂度和二叉树支持向量机（Binary Tree Support Vector Machine,BT-SVM）的心音分类识别方法。对心音进行经验模式分解（Empirical Mode Decomposition,EMD）,得到若干反映心音本体特征的平稳IMF分量;利用互相关系数准则对其筛选,计算所选IMF分量的复杂度值为信号的特征;将其组成特征向量输入到BT-SVM进行分类识别。临床数据仿真结果表明,该方法能有效提取心音特征,与传统识别方法相比,具有训练时间短,识别率高等优点。     

基于希尔伯特-黄变换的心音信号分析

人造心脏瓣膜心音信号是一种典型的非平稳随机信号,传统的傅里叶变换和小波分析的方法很难分析出其内在的特征,本文中采用希尔伯特-黄变换（HHT）将心音信号进行频谱分析,直观显示心音信号的频率分布,有助于对病人的心脏瓣膜进行监测,早期发现人工心脏瓣膜的病变。     

基于智能手机的心音监测系统的研究与设计

心音监测对心脏疾病的检测和预防有重要作用,研究并设计基于智能手机的心音监测系统。该系统由心音测量节点、智能手机节点、服务器端组成。测量节点实现心音信号的采集,智能手机节点接收测量节点的心音并将其经GPRS网络传递到服务器端,服务器端对心音进行远程的监测。其中智能手机节点是该系统的枢纽,其主要功能是分析处理心音：对心音进行小波去噪处理,通过LZ复杂度算法获取心音的特征指标,将特征指标用于心功能的分析评价及其异常预警。通过功能性测试表明该系统能稳定运行。     

基于ICA和小波变换的轴承故障特征提取

应用独立分量分析方法和小波变换分离轴承的振动信号，提取其状态特征。并对信号进行自相关预处理，突出信号的非高斯成分，较好地满足独立分量分析的前提条件，即源信号统计独立。采用基于负熵的快速独立分量分析（ICA）算法，成功地分离出了信号的一些独立成分。对ICA处理后的分量信号进行小波变换，完成信号检测，消噪，频带分析，以获取故障信号特征，确定故障的位置和强度。研究结果表明，独立分量分析方法和小波变换能提取明显的轴承故障信号特征。     

定量递归分析和近似熵在心音分类识别中的应用

为了有效利用心音信号的非线性特征信息对心音信号进行分类识别,提出一种基于定量递归分析和近似熵的心音特征提取方法.首先利用递归图对心音信号进行定性分析;然后,定量提取心音的非线性特征参数:递归率、确定率、近似熵构成特征矢量;最后将特征矢量输入二叉树支持向量机,对采集到的正常以及5类心脏瓣膜性心音信号进行分类识别.对于文中提取的非线性特征参数,通过统计学分析证明了其有效性.结果表明,该方法能有效识别心音信号.     

深度卷积神经网络在心音分类方法中的应用

通过分析心音信号对心脏早期的病理状态进行确诊具有重要的意义。提出了一种基于深度卷积神经网络的心音分类方法。将心音信号转化成具有时频特性的梅尔频谱系数（Mel Frequency Spectral Coefficient,MFSC）特征图,将其作为深度卷积神经网络模型的输入;利用深度卷积神经网络对MFSC特征图进行训练,引入中心损失函数建立最优的深度学习模型;测试阶段,先将心音信号转换成多张二维MFSC特征图,然后利用训练好的深度学习模型对其分类,最后利用多数表决原则判断心音信号的类别。针对人工标注的训练样本有限,导致模型训练正确率不高的问题,以心音的二维MFSC特征图为对象分别从时间域和频率域进行随机屏蔽处理进而扩充训练样本。实验结果表明,该方法在PASCAL心音数据集上进行测试,对正常、杂音、早搏三种心音的分类性能明显优于现有最好的方法。     

基于独立分量分析和BP网络的电子鼻模式识别

为了提高电子鼻对混合气体的识别率,针对气体传感器阵列的交叉敏感特性,探讨了在电子鼻系统中基于独立分量分析(ICA)算法与BP神经网络相结合进行模式识别的可行性。并对4个气体传感器组成的电子鼻对4种气体混合物所测得的原始数据进行处理,结果表明:ICA算法对数据进行有效预分类,减少了样本之间的相关性,将生成的新样本作为BP网络的输入,使网络结构简化,在保证一定正确率的前提下,大大提高网络的学习速度。利用该方法可以提高电子鼻识别混合气体的准确率。     

一种融合二维和三维卷积网络的两阶段冠状动脉分 割方法

心血管疾病是一种严重危害公众健康的重大疾病。与其他心血管疾病相比,冠心病是导致死亡的最主要原因,精确的冠状动脉分割对冠心病的治疗有重要意义。目前,深度学习已经广泛应用于医学影像领域,然而,像冠状动脉这样的小物体的分割仍然是一大挑战。针对冠状动脉精确分割的需求,该研究提出了一种融合二维和三维卷积网络的方案,利用骨架作为桥梁,结合二维和三维卷积网络,扩大了卷积网络的信息接受域。与其他深度学习方法相比,该方法在敏感度、Dice 系数、ROC 曲线下方的面积、豪斯多夫距离上均有一定程度的提升,且可以检测其他方法无法识别的冠状动脉,一定程度上解决了血管断连和血管缺失等问题。     

Solvation,transduction and independent component analysis for pattern recognition in SAW electronic nose

This paper revisits the fundamental basis for signal generation in polymer coated SAW vapor sensors and applies the independent component analysis (ICA) for feature extraction from the SAW sensor array data to explore whether the independent components could represent analyte-specific solvation parameters and whether they could form the feature vector for reliable pattern classification. Thermodynamic partitioning of analytes between vapor and polymer phases is treated as independent contributions from different solvation mechanisms, each associated with characteristic ‘environment swap’ energy. The overall equilibrium partition coefficient of an analyte is modeled as product of partial partition coefficients associated with different solvation mechanisms. The polymer films on SAW devices are treated to be acoustically thin. The theory of signal generation accounts for effects from both the mass as well as the viscoelastic loadings. It explains the signal amplification factor due to viscoelastic effects, and models the sensor signal to be proportional to the equilibrium partition coefficient. Thus, the logarithmic signal becomes a linear combination of the partial free energies associated with various solvation mechanisms. A linear-solvation-energy relationship (LSER) like factorization is assumed for the partial free energies where the latter are expressed as product of analyte and complimentary polymer associated solvation parameters. The problem of sensor array signal analysis is then treated as a blind source separation problem with the analyte solvation parameters being the independent sources, the polymer solvation parameters being the mixing weights and the log(signals) being the measured variables. The FastICA algorithm with Gram-Schmidt orthogonalization is applied to determine independent components. The principal component analysis (PCA) is done as pre-processing step for ICA. An experimental SAW sensor array data available in the literature [Rose-Pehrsson et al., Anal. Chem. 60 (1988) 2801–2811] is used to seek validation for our approach, and to examine the role of ICA in SAW sensor array signal processing. In brief, the paper establishes a direct relationship between the independent components and the analyte solvation parameters, and presents ICA as an effective method for feature extraction for pattern recognition in SAW electronic noses.     

基于Hilbert-Huang Transform的心音信号谱分析

心音信号是一种典型的非平稳信号,传统信号处理方法的应用受到很大限制.针对此本文提出了基于Hilbert-Huang Transform(HHT) 的心音信号的分析方法,对冠心病患者的心音信号进行了分析.通过把心音信号分解为内蕴模式函数,利用Hilbert变换建立了心音信号的时间-频率-能量三维Hilbert谱分布以及边界谱分布;Hilbert谱及其边界谱在时域以及频域以较高的分辨率表征了心音信号的时频变化特性,揭示了冠心病患者心音信号的病理特征;为冠心病的早期无损诊断奠定了坚实基础,临床实践中有较大的指导价值.     

几种神经网络在混合气体预测中的比较

气体传感器阵列的交叉敏感性严重影响气体传感器对混合气体的测量。用M atlab平台的神经网络工具箱,分别构建了BP,径向基(RBF)和模糊(FNN)神经网络,利用掺杂不同材料的4种SnO2气体传感器组成阵列,实现对甲醛、甲苯、丙酮和乙醇混合气体的体积分数预测。结果表明:FNN神经网络对混合气体体积分数预测的精度要高于其他2种网络。而且,结合PCA和ICA对数据样本进行预处理,有利于提高神经网络对体积分数预测的精度。     

PC based monitoring of human heart sounds

The paper presents an instrumentation system developed for monitoring of human heart sounds. A condenser microphone senses the heart sounds converting them into an equivalent electrical signal. The signal is suitably amplified and filtered in desired frequency band. This signal is converted to equivalent digital signal by an A/D conversion circuitry developed for the purpose. This digital signal is then fed to PC through the printer port (Syntronix). With the help of the supporting software, the signal for a specific duration is accessed and stored in the PC memory. This is further processed for its frequency contents. The system is simple and compact and does not require any external A/D card for the PC. It can be very useful for monitoring human heart sounds and interpreting the sounds on the basis of their time domain and frequency domain representation to diagnose heart disorders.     

基于体域网的个人健康监护系统设计

体域网(BSN,body sensor networks)是以人为中心,由分布在人体表、贴身衣物上,或身体内部检测人体生命体征的多个传感器节点以及个人智能终端组成的无线通信网络.介绍了一种基于ZigBee和蓝牙无线通信协议的个人健康监护体域网系统,该系统通过3G智能手机终端控制穿戴在人体上的传感器节点,实时采集人体血氧、心音、心电和血压等生命体征参数,并以无线通信方式依次传送至智能终端显示,进一步实现与社区医院或中心医院的远程数据交互.系统终端应用程序运行在Android操作系统下,界面设计友好,适用于所有Android操作系统的3G智能手机用户.     

Visual learning and classification of human epithelial type 2 cell images through spontaneous activity patterns

Identifying the presence of anti-nuclear antibody (ANA) in human epithelial type 2 (HEp-2) cells via the indirect immunofluorescence (IIF) protocol is commonly used to diagnose various connective tissue diseases in clinical pathology tests. As it is a labour and time intensive diagnostic process, several computer aided diagnostic (CAD) systems have been proposed. However, the existing CAD systems suffer from numerous shortcomings due to the selection of features, which is commonly based on expert experience. Such a choice of features may not work well when the CAD systems are retasked to another dataset. To address this, in our previous work, we proposed a novel approach that learns a set of filters from HEp-2 cell images. It is inspired by the receptive fields in the mammalian's vision system, since the receptive fields can be thought as a set of filters for similar shapes. We obtain robust filters for HEp-2 cell classification by employing the independent component analysis (ICA) framework. Although, this approach may be held back due to one particular problem; ICA learning requires a sufficiently large volume of training data which is not always available. In this paper, we demonstrate a biologically inspired solution to address this issue via the use of spontaneous activity patterns (SAP). The spontaneous activity patterns, which are related to the spontaneous neural activities initialised by the chemical release in the brain, are found as the typical stimuli for the visual cell development of newborn animals. In the classification system for HEp-2 cells, we propose to model SAP as a set of small image patches containing randomly positioned Gaussian spots. The SAP image patches are generated and mixed with the training images in order to learn filters via the ICA framework. The obtained filters are adopted to extract the set of responses from a HEp-2 cell image. We then employ regions from this set of responses and stack them into “cubic regions”, and apply a classification based on the correlation information of the features. We show that applying the additional SAP leads to a better classification performance on HEp-2 cell images compared to using only the existing patterns for training ICA filters. The improvement on classification is particularly significant when there are not enough specimen images available in the training set, as SAP adds more variations to the existing data that makes the learned ICA model more robust. We show that the proposed approach consistently outperforms three recently proposed CAD systems on two publicly available datasets: ICPR HEp-2 contest and SNPHEp-2.     

气体传感器阵列温度漂移抑制的研究

半导体气体传感器存在漂移问题,温度变化对漂移的影响尤为明显.在气体传感器阵列中,可以加入温度、湿度等传感器,监测其工作环境.实验系统采用恒温箱设定一组温度,制备气体样本20例(两种浓度样本各10例),采集传感器对样本的响应;通过人工神经网络来识别样本;当有误判发生时,在原网络中引入温度传感器的响应值,消除了误判,在一定程度上抑制了漂移,改善了网络性能,验证了该温度漂移抑制方法的可行性.     

Integration of independent component analysis and neural networks for ECG beat classification

In this paper, we propose a scheme to integrate independent component analysis (ICA) and neural networks for electrocardiogram (ECG) beat classification. The ICA is used to decompose ECG signals into weighted sum of basic components that are statistically mutual independent. The projections on these components, together with the RR interval, then constitute a feature vector for the following classifier. Two neural networks, including a probabilistic neural network (PNN) and a back-propagation neural network (BPNN), are employed as classifiers. ECG samples attributing to eight different beat types were sampled from the MIT-BIH arrhythmia database for experiments. The results show high classification accuracy of over 98% with either of the two classifiers. Between them, the PNN shows a slightly better performance than BPNN in terms of accuracy and robustness to the number of ICA-bases. The impressive results prove that the integration of independent component analysis and neural networks, especially PNN, is a promising scheme for the computer-aided diagnosis of heart diseases based on ECG.