DNA算法优化BP网络用于心律失常的识别

背景:目前对心律失常的诊断大多是由医生人工完成,费时费力,诊断结果依赖于医生的个人业务水平和责任心。心律失常的自动识别对于心脏病患者的救护和早期治疗具有非常重要的意义。目的:实现临床心律失常的自动识别和诊断。方法:首先从心电图中动态提取完整心律失常心拍形态,并采用离散余弦变换和反变换压缩数据;然后设计用于心律失常识别的BP神经网络,并用DNA算法优化该BP网络;最后用MIT/BIH心电数据库中心电图数据对DNA-BP网络进行检验。结果与结论:对于5种心拍类型,包括正常、左束支阻滞、右束支阻滞、心室跳脱心搏及Paced心搏,利用DNA-BP网络进行分类,实验达到了很好的识别效果,平均识别正确率达到99%。     

植入式除颤器腔内心电图的分析方法

植入式除颤器（ICD）自1980年首次应用于临床后,已经成为用于治疗恶性心律失常、减少患者死亡率的一种重要的医疗手段。如何正确对心律失常进行分类识别、避免引发不适当电击仍然是植入式除颤器的关键问题。通过分析腔内心电图来识别心律失常的方法主要有：心率阈值、突发性、稳定性、形态学分析及时频分析,人工神经网络以及协方差分析等。其中节律检测识别算法和波形相关性分析是ICD中常用的检测方法,也是其他检测方法的基础。随着技术的发展,采用DSP芯片作为处理器的人工神经网络等方法也有望在今后的ICD中得到应用。     

基于卷积神经网络特征提取与融合的心律失常分类

本研究提出一种新的心律失常自动分类方法,辅助医生诊治心律失常。通过构建卷积神经网络对心电信号以及QRS波群的小波分量进行特征提取,将网络提取到的心电信号特征和小波特征与人工提取的RR间期特征,输入到全连接层进行融合,在输出层使用softmax函数对心拍进行分类。使用MIT-BIH心律失常数据库中的MILL导联数据对网络进行训练和测试。经测试,该方法的总体分类准确度达98.12%,平均灵敏度为87.32%,平均阳性预测值为90.37%。该方法能够快速识别不同类型的心律失常,对于计算机辅助诊断心律失常的应用具有一定的参考价值。     

心律失常辅助诊断系统的研制

目的:研制心律失常辅助诊断系统,以减少医生的工作量,并提高其对心电信号诊断的准确性。方法:首先利用小波变换理论建立滤波和波形识别算法,提取出有效的特征参数;然后利用粗糙集理论约简特征参数并根据相应的分类决策规则,利用分支逻辑法对波形进行识别分类;最后利用模糊神经网络理论得出异常心拍的隶属度。结果:实现了滤波、波形识别、诊断分类等主要模块,形成了一个完整的系统。结论:该系统能识别十九种心律失常并得出异常心拍的隶属度和位置信息,对医生的诊断有良好的辅助作用。     

心律失常波序识别的随机Patri网络模型

许多心律失常可以只根据ECG中各种波形的时序关系来进行诊断,然而要进行心律失常的自动识别就要将各种波形的模式用数学方法表达。这种数学表达一方面要很复杂以便包含更多的波形模式,另一方面又要尽量简单化以使这种分类算法能够实现。本文介绍了一种根据ECG来模拟心脏节律的方法。文中介绍了这种能反映各种不同心律失常的P、R波时序关系的动态模型。它是把心脏作为一个聚合体,用     

心律失常分析方法和算法探讨

本文讨论了心律分析的方法和算法设计、提出了逻辑滤波作为信号的预处理和波峰的预识别,以及QRS波群逐峰定位算法、于漂移程度的基线拟合算法、LOT变换和逻辑推断的任意区段P波知别算法及P波节律分析法方法等。这些算法具有速度快、识别效果好的特点,特别适用于大数据量的心律分析。并且基于预波峰识别的一系列算法适合于复杂类的心律失常分析,是心律失常自动分析方法探索的一个方向。     

阈下条件电刺激心室抑制作用空间限制性的实验研究

临床上,已利用阈下条件刺激心室抑制作用的空间限制性对室性心律失常异位激动起源点进行定位。为了解这种方法定位的精确性,本工作采用各种观察阈下条件刺激的心室抑制作用及其空间限制性的方法,并首次设计了阈下条件刺激心室起搏节律抑制模型,进行了详细研究。结果表明,阈下条件刺激的心室抑制作用具有很强的空间限制性,这一特性可用来对室性心律失常异位激动起源点进行精确定位。     

无线心电遥测监护在恶性心律失常诊断中的价值

目的 探讨无线心电遥测监护在恶性心律失常诊断中的价值.方法 系统复习近四年2100例患者的心电监护资料,并根据心律失常的类型而探讨无线心电遥测监护对恶性心律失常诊断的价值.结果 共检出各种心律失常1785例,其中包括恶性窦性心律失常,恶性房性心律失常,恶性室性心律失常等.结论 无线心电遥测监护对恶性心律失常的诊断具有极其重要的意义,必将在将来具有进一步的发展.     

基于焦点损失函数的嵌套长短时记忆网络心电信号分类研究EI北大核心CSCD

心电图(ECG)可直观地反映人体心脏生理电活动,在心律失常检测与分类领域中具有重要意义。针对ECG数据中类别不平衡对心律失常分类带来的消极作用,本文提出一种用于不平衡ECG信号分类的嵌套长短时记忆网络(NLSTM)模型。搭建NLSTM学习并记忆复杂信号中的时序特征,利用焦点损失函数(focal loss)降低易识别样本的权重;然后采用残差注意力机制(residual attention mechanism),根据各类别特征重要性修改已分配权值,解决样本不平衡问题;再采用合成过采样技术算法(SMOTE)对麻省理工学院与贝斯以色列医院心律失常(MIT-BIH-AR)数据库进行简单的人工过采样处理,进一步增加模型的分类准确率,最终应用MIT-BIHAR数据库对上述算法进行实验验证。实验结果表明,所提方法能有效地解决ECG信号中样本不平衡、特征不突出的问题,模型的总体准确率达到98.34%,较大地提升对少数类样本的识别和分类效果,为心律失常辅助诊断提供可行的新方法。     

基于QRS波群的心律失常辅助诊断模型研究

基于对QRS波群的特征变量提取,利用减法聚类和自适应模糊神经网络构建心律失常辅助诊断模型,分析不同训练数据集对模型测试结果的影响.实验结果表明,该模型能准确识别不同类型的QRs波群,使用不同训练数据集对诊断结果存在影响,为进一步实现更复杂的心律失常辅助诊断模型提供方法.     

A Construction Method of Personalized ECG Template and Its Application in Premature Ventricular Contraction Recognition for ECG Mobile Phones

Premature ventricular contraction(PVC) is the most frequent arrhythmia encountered in clinical practice. PVC may occur in health subjects, which is not imminently life-threatening but may require therapies to prevent further problems. So,the timely PVC recognition becomes very important for the analysis of electrocardiogram(ECG), especially for the remote ECG monitoring using mobile phones. In this paper,a construction method of personalized ECG template and a PVC recognition method based on template matching were studied. Firstly, we selected 43 ECG recordings from the MIT-BIH arrhythmia database. All recordings were divided into two datasets(DS1for training and DS2 for testing) and each dataset approximately contained the same proportion of PVC beats. Subsequently, for each recording(30 min) in DS1, the first5 min recordings were used to construct the personalized ECG template and the last25 min recordings were used for the R-wave peaks detection and PVC recognition,where the template matching method were used. The validity of the proposed methods was tested using DS2. The results showed that: 1) high beat detection accuracy was achieved for both PVC beats and non-PVC beats; 2) the sensitivity and specificity of PVC recognition were 99.11% and 99.96% for the first 5 min recordings respectively,99.17% and 99.43% for the last 25 min recordings respectively. All the proposed methods can be real-time performed, which show a promising prospect for the application of ECG mobile phones.     

基于支持向量机的室性早搏检测

心电信号分类是自动心电监护设备的基础。支持向量机 (SVM)在分类和模式识别方面展现出卓越的性能。本研究将支持向量机应用于心电信号室性早搏 (PVC)的检测。根据室性早搏的特点 ,从 ML II导联中提取心率、形态心及小波域能量 3大类共 9个特征。并使用 MIT- BIH的 Arrhythmia数据库的数据 ,根据 AAMI建议要求 ,对采用不同核函数的支持向量机的性能作了比较。     

HRV的最大李雅普诺夫指数估计

对心率变异性(HRV)进行了研究,比较了心率正常者与心率失常者HRV之间的最大李雅普诺夫指数上的差别。人在正常状态和病理状态下的HRV信号最大李雅普诺夫指数是不同的,当出现病理心血管事件时,指数α减少,因此李雅普诺夫指数可作为人体是否异常或处于何种异常状态的特征刻画指标,本文心率正常者HRV信号的最大李雅普诺夫指数为0.45907,心率不齐者的最大李雅普诺夫指数是0.41472。它们均为混沌信号,但是处于心率不齐状态的节律混沌程度明显比处于心率正常状态的节律混沌程度低。     

Familial cardiomyopathy: Autosomally, dominantiy inherited congestive cardiomyopathy with two cases of septal hypertrophy in one family

A family with inherited congestive cardiomyopathy is presented. The diagnosis is based on clinical, morphological and laboratory evaluations. The first observed sign of the disease is arrhythmia and/or conduction defects. The onset of symptoms of pump failure is in adult life, and affected persons die within several years. Three persons have died suddenly. Septal hypertrophy was present in two affected persons. The mode of transmission is probably autosomal dominant. The recognition of arrhythmia as an early sign of the disease offers the opportunity of an early diagnosis.     

Classification enhancible grey relational analysis for cardiac arrhythmias discrimination

This paper proposes a method for electrocardiogram (ECG) heartbeat recognition using classification enhancible grey relational analysis (GRA). The ECG beat recognition can be divided into a sequence of stages, starting with feature extraction and then according to characteristics to identify the cardiac arrhythmias including the supraventricular ectopic beat, bundle branch ectopic beat, and ventricular ectopic beat. Gaussian wavelets are used to enhance the features from each heartbeat, and GRA performs the recognition tasks. With the MIT-BIH arrhythmia database, the experimental results demonstrate the efficiency of the proposed non-invasive method. Compared with artificial neural network, the test results also show high accuracy, good adaptability, and faster processing time for the detection of heartbeat signals.     

Support vector machine-based arrhythmia classification using reduced features of heart rate variability signal

OBJECTIVE: This paper presents an effective cardiac arrhythmia classification algorithm using the heart rate variability (HRV) signal. The proposed algorithm is based on the generalized discriminant analysis (GDA) feature reduction scheme and the support vector machine (SVM) classifier. METHODOLOGY: Initially 15 different features are extracted from the input HRV signal by means of linear and nonlinear methods. These features are then reduced to only five features by the GDA technique. This not only reduces the number of the input features but also increases the classification accuracy by selecting most discriminating features. Finally, the SVM combined with the one-against-all strategy is used to classify the HRV signals. RESULTS: The proposed GDA- and SVM-based cardiac arrhythmia classification algorithm is applied to input HRV signals, obtained from the MIT-BIH arrhythmia database, to discriminate six different types of cardiac arrhythmia. In particular, the HRV signals representing the six different types of arrhythmia classes including normal sinus rhythm, premature ventricular contraction, atrial fibrillation, sick sinus syndrome, ventricular fibrillation and 2 degrees heart block are classified with an accuracy of 98.94%, 98.96%, 98.53%, 98.51%, 100% and 100%, respectively, which are better than any other previously reported results. CONCLUSION: An effective cardiac arrhythmia classification algorithm is presented. A main advantage of the proposed algorithm, compared to the approaches which use the ECG signal itself is the fact that it is completely based on the HRV (R-R interval) signal which can be extracted from even a very noisy ECG signal with a relatively high accuracy. Moreover, the usage of the HRV signal leads to an effective reduction of the processing time, which provides an online arrhythmia classification system. A main drawback of the proposed algorithm is however that some arrhythmia types such as left bundle branch block and right bundle branch block beats cannot be detected using only the features extracted from the HRV signal.     

HRV信号关联维计算中参数选取的研究

健康和有疾病的心率变异性（HRV）参数有明显差异,计算关联维是识别这种差异的一种重要手段。用传统的G—P算法计算关联维时,嵌入维数m、延迟时间τ、及序列长度N等参数的选取会对最终计算结果有很大影响。本文从理论和实验方面论述了如何选取这些参数以获得正确的结果,并且将其应用于正常组和心率不齐疾病组进行对照,结果显示关联维可以有效地表征由于疾病对于心脏节律造成的影响。     

Detection of ventricular fibrillation using empirical mode decomposition and Bayes decision theory

Ventricular fibrillation (VF) is the most serious variety of arrhythmia which requires quick and accurate detection to save lives. In this paper, we propose an empirical mode decomposition (EMD) based algorithm for VF detection. The intrinsic mode functions (IMFs) of VF are orthogonal whereas the lower order IMFs of normal sinus rhythm (NSR) are not. The orthogonality indices derived from the first three consecutive intrinsic mode functions (IMFs) of NSR and VF are used for their discrimination. The proposed technique is applied to the MIT-BIH arrhythmia database. The accuracy of detection of VF is 99.70% for a window length of 3 s. This early estimate of VF may be useful in emergency cases where defibrillators are to be applied. Comparative results with the existing methods in terms of quality parameters and integrated receiver operating characteristic (IROC) are presented.     

基于总体局部均值分解方法的心律失常特征提取与分类

针对心电信号自动分类技术中的特征提取,提出一种新的特征提取方法—总体局部均值分解（ELMD）方法。该方法首先对心电信号加入不同的高斯白噪声,然后进行局部均值分解得到若干乘积函数（PF）分量,求取多次分解后的PF分量均值。多次加入噪声及分量平均的过程可以克服基本局部均值分解方法存在的模态混叠问题。选取较优的前4个PF分量进行特征计算,将得到的特征向量矩阵送入支持向量机对正常心电信号和4种常见的心律失常信号进行分类。从MIT-BIH心律失常数据库的分类结果来看,ELMD总体分类准确率达到99.61%,高于一般方法,证明了ELMD方法的有效性。     

Changes in the distribution of ventricular ectopic beats in long-term electrocardiograms

Only simple methods have been used to assess antiarrhythmic and proarrhythmic effects when comparing baseline and on-therapy Holter recordings taken from the same patient. The paper suggests a new method for definition of these effects based upon comparisons of statistical distributions of ectopic beats. The method is based on multiple random sampling of the recordings and on application of the Smirnov test to compare the samplings. The results of these multiple comparisons are subsequently evaluated using the chi-squared test. An analysis is reported of Holter recordings made on seven patients suffering from ventricular arrhythmia but with anatomically normal heart. In each patient, one baseline recording and one recording on each of three different drugs were made. The results show that the definition of proarrhythmic effects can be partly addressed in a precise mathematical way. The method can also detect a significant change in the character of arrhythmia which can neither be classified as antiarrhythmic nor as proarrhythmic.