高介神经网络用于心电信号的识别

心电图（ＥＣＧ）的自动识别用分类长期以来一直是较难解决问题，特别是当截取ＥＣＧ信号的起点不同（相当于信号发生平移）及有基线漂移（相当于信号发生旋转）时，在不作预处理情况下要求仍能正确识别及分类，即要求识别过程具有平移不变性和旋转不变性，更是困扰生物医学工程工作者的困难问题之一。作者采用高阶神经网络（二阶）、对五类（正常类，高Ｒ波类、高Ｔ波类、Ｔ波倒置类及心律不齐类）具有平移及旋转的ＥＣＧ信号进行分     

Holter系统的技术要求及神经网络方法

为了长时间、准确的记录分分析体表ＥＣＧ，从各个方面满足基础研究、临床、康复以及特殊生理需要，对Ｈｏｌｔｅｒ系统提出了愈来愈高的要求。作者从技术实现的角度讨论了Ｈｏｌｔｅｒ的主要技术要求，包括记录部分和分析部分；还介绍了几种用神经网络模型满足Ｈｏｌｔｅｒ系统技术要求的原理，包括用多层感知器实现数据压缩及分类，用高阶神经网络实现ＥＣＧ的分类，以及用ＡＲＴ模型识别ＥＣＧ中的Ｐ波与Ｔ波。     

高速枪弹压力波调控内皮细胞ET—1的信号通路

目的和方法：以高速枪弹压力波致伤培养细胞为实验模型，以特异的Ｇ蛋白和肌醇磷脂代谢抑制剂作用培养内皮细胞，力图从分子创伤学角度探讨压力波调控ＥＴ－１的信号通路。结果：磷脂酶Ｃ抑制剂ｎｅｏｍｙｃｉｎ作用内皮细胞，明显消除压力波对ＥＴ－１释放及表达的刺激作用。Ｇ蛋白抑制剂ＧＤＰ－β－Ｓ则无明显作用。蛋白激酶Ｃ抑制剂ｓｌａｕｒｏｓｐｏｒｉｎｅ对压力波作用后ＥＴ－１基因表达增强有较强的抑制作用。结论：高速枪弹压力波对内皮细胞ＥＴ－１的调控可能通过某种直接途径（而非Ｇ蛋白介导）激活肌醇磷脂代谢，从而生成ＩＰ３和二酰基甘油。ＩＰ３可能参与早期ＥＴ－１释放，二酰基甘油对ＰＫＣ的激活参与了ＥＴ－１基因表达的调控。     

降钙素基因相关肽对内皮素释放的影响

本实验在大鼠的整体及离体血管条上观察了降钙素基因相关肽（ＣＧＲＰ）对内皮素（ＥＴ）释放的影响。结果表明：ＣＧＲＰ静脉注射（５μｇ／ｋｇ）能明显降低大鼠内毒素血症时血浆ＥＴ含量，ＣＧＲＰ孵育（１０＾－８ｍｏｌ／Ｌ）离体的大鼠主动脉血管条能有效地抑制凝血酶（ｔｈｒｏｍｂｉｎ）引起的ＥＴ释放，但是ＣＧＲＰ静注及孵育不影响血浆ＥＴ的基础含量及离体血管条ＥＴ的基础释放。结果提示ＣＧＲＰ能抑制病理条件下ＥＴ的     

抗鹭山病毒单克隆抗体的研制及应用

本文报告应用淋巴细胞杂交瘤技术，共获得５株能持续、稳定分泌抗鹭山（ＳＡＧ）病毒单克隆抗体的杂交瘤细胞系，间接免疫荧光法证实，两株ＭｃＡｂ具有病毒种的特异性，其它３株具有亚组特异性。免疫印迹分析发现，前者仅能识别ＳＡＧ病毒Ｃ蛋白上的表位，后者能识别ＳＡＧ病毒的三个结构蛋白（Ｅ１、Ｅ２和Ｃ），并与ＧＥＴ、ＭＡＹ及ＲＲ鹭甲病毒的Ｃ蛋白有明显的交叉反应。     

抗鹭山病毒单克隆抗体的研制及应用

本文报告应用淋巴细胞杂交瘤技术，共获得５株能持续、稳定分泌抗鹭山（ＳＡＧ）病毒单克隆抗体的杂交瘤细胞系，间接免疫荧光证实，两株ＭｃＡｂ具有病毒种的特异性，其它３株具有亚组特异性。免疫印迹分析发现，前者仅能识别ＳＡＧ病毒Ｃ蛋白上表位，后者能识别ＳＡＧ病毒的三个结构蛋（Ｅ１、Ｅ２和Ｃ），并与ＧＥＴ、ＭＡＹ及ＲＲ鹭甲病毒的Ｃ蛋白有明显的交叉反应。     

使用神经网络检测心电图(ECG)“左心室劳损”

使用神经网络检测心电图（ＥＣＧ）“左心室劳损”［英］／ＢｅｖｉｎｅＢ…／／ＭｅｄＢｉｏｌＥｎｇＣｏｍｐｕｔ．－１９９３，Ｊｕｌｙ．－３４３本文研究了利用人工神经网络对ＥＣＧ的ＳＴ—Ｔ段异常的分类。我们从１０５例ＥＧＣ中选出３５６个侧导联的数据组成训练...     

Eysenck人格维度与心向量图横面T环顺钟向运行及发生心脏 …

目的：探讨人格维度心向量图（ＶＣＧ）横面Ｔ环顺钟向运行（顺转）及发生心脏神经官能症（ＣＮ）的关系，方法：采用Ｅｙｓｅｎｃｋ个性问卷评定了２２４例ＶＣＧ横向Ｔ环顺转及发生ＣＮ者的气质特征。结果：ＶＣＧ横面Ｔ环顺转及发生ＣＮ与情绪性质质特征相关显著，多血胆汁是Ｔ环顺转及发生ＣＮ者的多见气质类型。结论：气质与ＶＣＧ横面Ｔ环顺转及发生ＣＮ显著相关。     

用移动平均幅值微分算法检测胎儿R波

用移动平均幅值微分算法检测胎儿Ｒ波［英］／Ｐａｒｋ　ＹＣ…／／ＩＥＥＥＴｒａｎｓＢＭＥ．－１９９２，３９（８）．－８６８胎儿ＥＣＧ信号处理问题包括从腹导ＥＣＧ信号中去除母体ＱＲＳ复合波增强胎儿ＱＲＳ复合波，检测胎儿Ｒ波计算胎儿心率。ＡＭＤＦ（ａｖｅｒ...     

22例继发性癫痫皮层电图与脑电图的关系

对我院１９９２～１９９４年２２名继发性癫痫患者的皮层电图（ＥＣｏＧ）与普通脑电图（ＥＥＧ）的关系，结合ＣＴ及病理所见作了对比。本组３种类型癫痫患者的皮层痫灶主要集中在额中、颞中、前颞，三者占总数的８５％。ＥＣｏＧ上相邻３个电极出现棘波发作者，在ＥＥＧ上可出现相应的棘、尖波灶，反之则不然。ＥＥＧ多形性δ波灶提示患者脑器质性病变，应进一步作影象检查。ＥＣｏＧ可在病灶周围找到皮层棘波灶。     

心电图自动诊断系统的研制

本文利用数字信号处理和波形识别技术，对心电图的ＱＲＳ波、Ｐ波、Ｔ波的时段进行了计算机的自动检测并对室性期前收缩等１５种异常心电进行自动诊断。打印输出相应的诊断报告。为了检验本系统的稳定性和可信度，本文还利用美国麻省理工学院的ＭＩＴ－ＢＩＨ数据库对本文所使用的方法进行了检测，取得了较好的效果。     

基于提升小波变换和多种策略的QRS波检测算法

目的：为了提高计算机处理心电信号的速率和精度,提出了一种基于提升小波变换,结合多种策略的QRS波检测算法。方法：首先采用基于阀值的提升小波去噪方法去除心电信号中的高频白噪声和低频基线漂移;再对处理后的心电信号进行提升小波分解,得出各层逼近信号和细节信号,在第3尺度上采用模极大值阀值法对R波进行检测．找出备选的R波,同时采用几何的方法定位Q波和S波及QRS波起点和终点;最后采用补偿法、波宽法及QRS波时长法对QRS波群进行纠正。结果：本文算法在时域心电图上实现了QRS波的准确定位．提取了心电图的QRS波段。通过MIT—BIH数据库验证,本算法具有很好的表现。结论：实验结果表明,相比传统的算法,本文采用的提升小波和多种策略的检测算法．能有效的检测QRS波,为心电信号的自动识别奠定了基础。     

Effects of noise and filtering on SVD-based morphological parameters of the T wave in the ECG

Singular value decomposition (SVD) based electrocardiogram (ECG) morphology analysis is a novel method in the assessment of subtle abnormalities in the T wave morphology of 12-lead ECG. As various types of noise contaminate the ECG signal and create a bias for the morphological analyses, this study was designed to estimate the effects of noise on the SVD method in an experimental setup. Ideal signals were generated by filtering real ECG signals several times with the Savitzky-Golay filter. Random and real noise samples were superimposed on the ideal signals. The noisy signals were filtered with a power line interference filter combined with the Savitzky-Golay or the wavelet filter. Results show that noise increased both the dipolar and non-dipolar components significantly unless filtering was applied. R-TWR (relative T wave residuum) and A-TWR (absolute T wave residuum) were four to eight times higher in noisy signals. The experiments with patient data demonstrated that certain types of noise may even lead to erroneous classification of patients. Filtering brings the median values closer to the correct ones and decreases significantly the variance of the values of parameters.     

Effects of noise and filtering on SVD-based morphological parameters of the T wave in the ECG

Singular value decomposition (SVD) based electrocardiogram (ECG) morphology analysis is a novel method in the assessment of subtle abnormalities in the T wave morphology of 12-lead ECG. As various types of noise contaminate the ECG signal and create a bias for the morphological analyses, this study was designed to estimate the effects of noise on the SVD method in an experimental setup. Ideal signals were generated by filtering real ECG signals several times with the Savitzky-Golay filter. Random and real noise samples were superimposed on the ideal signals. The noisy signals were filtered with a power line interference filter combined with the Savitzky-Golay or the wavelet filter. Results show that noise increased both the dipolar and non-dipolar components significantly unless filtering was applied. R-TWR (relative T wave residuum) and A-TWR (absolute T wave residuum) were four to eight times higher in noisy signals. The experiments with patient data demonstrated that certain types of noise may even lead to erroneous classification of patients. Filtering brings the median values closer to the correct ones and decreases significantly the variance of the values of parameters.     

正交小波变换的快速算法在心电QRS波检测中的应用

目的：研究基于小波变换的心电QRS波检测的准确率、抗干扰性和实时性，论证其在实际工程应用中的可行性。方法：作者在比较了不同小波基的检测准确率之后，采用一种基于三次B样条小波变换的心电QRS波检测算法，利用离散正交二进小波的快速算法-Mallat算法进行分解滤波，再利用小波变换与信号奇异点的关系，在2^3尺度下识别R波峰值，在2^1尺度上检测QRS波的起点和终点，QRS波的起点和终点对应于小波变换的一对符号相反的模极大值，R波的峰点对应于介于这对模极大值之间的小波变换过零点，并用美国MIT／BIH心电标准数据库分析该算法的准确率、抗干扰性和实时性。结果：该方法具有比较理想的检测准确率，在99％以上；对肌电、工频、基漂等常见的心电信号干扰有较好的容限度，即使心电序列伴有严重的基漂和高频、工频、肌电等干扰，也不影响QRS波的检测；此外，三次B样条小波基的滤波器个数少，提高了运算速度，采样11．4s的数据进行分析，耗时为0．2s～0.3S，实时效果较明显。结论：可以满足实际工程应用的需要。     

面向家庭健康监护的心电监护系统设计

目的研制一款面向家庭的心电监护系统。方法以基于ARM920T的s3c2440为核心,控制心电信号采集,并结合嵌入式软件技术,实时显示、分析和记录心电信号,对患者进行监护。算法部分采用适用于嵌入式系统的动态差分阈值法检测QRS波波群。结果该监护系统能实时、动态显示心电波形,并可以识别4种心率异常,能较好地反映和分析患者的心电活动状况。结论该心电监护系统操作简便,运行稳定,能够满足一般家庭需求,具有良好的应用前景。     

Classification of ECG signals using LDA with factor analysis method as feature reduction technique

The analysis of ECG signal, especially the QRS complex as the most characteristic wave in ECG, is a widely accepted approach to study and to classify cardiac dysfunctions. In this paper, first wavelet coefficients calculated for QRS complex are taken as features. Next, factor analysis procedures without rotation and with orthogonal rotation (varimax, equimax and quartimax) are used for feature reduction. The procedure uses the ‘Principal Component Method’ to estimate component loadings. Further, classification has been done with a LDA classifier. The MIT-BIH arrhythmia database is used and five types of beats (normal, PVC, paced, LBBB and RBBB) are considered for analysis. Accuracy, sensitivity and positive predictivity are performance parameters used for comparing performance of feature reduction techniques. Results demonstrate that the equimax rotation method yields maximum average accuracy of 99.056% for unknown data sets among other used methods.     

用小波变换结合神经网络检测ECG信号的P波

通过小波变换对ＥＧＣ信号进行分解,然后采用神经网络检测ＥＣＧ信号的Ｐ波,该方法作为一种辅助检测手段,效果良好。将其用于心率变异性分析具有重要意义。     

Using simulated noise to define optimal QT intervals for computer analysis of ambulatory ECG

The ambulatory electrocardiogram (ECG) is an important medical tool, not only for diagnosis of adverse cardiac events, but also to predict the risk of such events occurring. The 24-hour ambulatory ECG has certain problems and drawbacks because the signal is corrupted by noise from various sources and also several other conditions which may alter the ECG morphology. We have developed a Windows based program for the computer analysis of ambulatory ECG which attempts to address these problems. The software includes options for importing ECG data, different methods of waveform analysis, data-viewing, and exporting the extracted time series. In addition, the modular structure allows for flexible maintenance and expansion of the software. The ECG was recorded using a Holter device and oversampled to enhance the fidelity of the low sampling rate of the ambulatory ECG. The influence of different sampling rates on the interval variability were studied. The noise sensitivity of the implemented algorithm was tested with several types of simulated noise and the precision of the interval measurement was reported with SD values. Our simulations showed that, in most of the cases, defining the end of QT interval at the maximum of the T wave gave the most precise measurement. The definition of the onset of the ventricular repolarization duration is most precisely made on the maximum or descending maximal slope of the R wave. We also analyzed some examples of time series from patients using power spectrum estimates in order to validate the low level QT interval variability.     

The comparison of different feed forward neural network architectures for ECG signal diagnosis

The electrocardiograms (ECGs) record the electrical activity of the heart and are used to diagnose many heart disorders. This paper proposes a two-stage feed forward neural network for ECG signal classification. The research is aimed at the design of an intelligent ECG diagnosis tool that can recognise heart abnormalities while reducing the complexity, cost, and response time of the system. A number of neural network architectures are designed and compared for their ability to classify six different heart conditions. Two network architectures based on one stage and two stage feed forward neural networks are chosen for this investigation. The training and testing ECG signals are obtained from MIT-BIH database. The network inputs are comprised of 12 ECG features and 13 compressed components of each heart beat signal. The performance of the different modules as well as the efficiency of the whole system is presented. Among different architectures, a proposed multi-stage network named NET_BST possesses the highest recognition rate of around 93%. Therefore, this network proves to be a suitable candidate in ECG signal diagnosis systems.