二维主元分析法在心肌梗死心电信号检测中的应用

利用基于联合能量百分比搜索的二维主元分析法对12导高分辨率心电信号（ECG）进行全局特征提取和分类检测研究。所用数据取自PTB诊断数据库,包括健康状态ECG,早期心肌梗死（MI）ECG,急性期MI ECG,恢复期 MI ECG。结果表明,所用的方法能有效地融合12导ECG信号及其高频分量中的细微结构信息,与常规主元分析法相比,其平均分类检测精度可提高10.43%,与常规二维主元分析法相比,能得到维数更低的特征表示,并可获得99.46 %的平均分类检测精度。     

基于多特征分支卷积神经网络的心电图分类算法

我国心血管疾病发病率、病死率呈逐年上升趋势。但由于心电图数据规模大且繁杂,临床医护人员在心电图筛查时,工作负担大且容易出现误诊或者漏诊的情况。基于此,利用CPSC-2018 12导联数据,提出了一种基于多特征分支卷积神经网络的多导联心电信号的智能分类与分析。首先,将CPSC-2018 12导联数据分为9个类别,基于12导联推导出8导联心电信号并分别提取局部特征。然后,通过双向GRU编码和注意力机制计算出不同类别的注意力权重向量,并将特征信息串联融合成特征向量,从而实现多导联心电图分类。实验结果表明：在验证集上取得了较好的分类效果,正常类别的F1值达到81.2%,平均F1值达到84.2%。特别地,在识别房颤（AF）和右束支传导阻滞（RBBB）这两类别心律失常时F1值分别达到95.1%和93.1%。     

多参数监护仪检定中常见故障分析及排除

各大医院在用多参数监护仪(以下简称"监护仪")品牌、型号繁多,在操作和使用上会有很大的变化,给计量检定工作带来困难。本文主要介绍笔者在多年检定工作中总结的监护仪常见故障分析及排除方法。一、ECG(心电)无波形输出打开监护仪,将心电导联线接在检定仪上,由检     

基于ECG和HRV特征提取的情绪状态分析

心电(ECG)和心率变异性(HRV)是心脏活动的直接响应,与自主神经系统有着密切关系,不同情绪状态有特定的自主神经反应模式,因而ECG和HRV可以反映人体情绪变化.本文利用Augsburg大学建立的情感生理数据库,选取音乐情感诱发下的心电信号,对其进行滤波处理去除噪声和干扰;采用小波变换提取心电信号小波系数的模极值和过零点,得到心电R波信号,计算其一阶差分,获得4种情绪状态下的心率变异性特征信号;提取情绪诱发的心电和心率变异性信号的时域、频域和小波熵时频特征,并对特征进行差异统计检验和情绪状态分析.仿真结果表明,HRV特征相比ECG特征可以更有效地反映人体的情绪状态,小波熵特征和HRV频谱总功率特征对于不同情绪状态的差异性更显著,该研究可为机器情绪识别提供借鉴.     

基于深度学习算法的正负性情绪识别研究

为探究感知正性情绪(高兴、惊奇、自豪、感动)和负性情绪(愤怒、悲伤、恐惧、厌恶)时,人体生理信号特征变化,并依此进行情感识别,设计了视频诱发情感的实验范式.利用多导生理仪同步采集人体脉搏和心电两种生理信号;采用中值滤波和小波去噪方法消除测量中的基线漂移和噪声干扰;通过差分阈值法进行峰值检测,提取心电和脉搏波特征;设计栈式自编码深度学习算法,利用神经网络分类器实现正负性情绪分类识别.基于心电信号特征或脉搏信号特征,4种正性情绪的平均识别正确率分别为83.16%和81.66%,4种负性情绪的平均识别率分别为90.33%和86.33%,4种正负性混合情绪的平均识别率分别达到87.86%和85.28%.结果表明:采用栈式自编码深度学习算法,基于脉搏和心电生理信号特征,均可以有效识别正负性情绪,并且心电特征相比脉搏特征在正负性情绪识别方面更优越,该方法可以应用于情感机器人的研究中.     

基于眼动特征及ECG的轨道交通驾驶员疲劳分析与识别

目的分析城市轨道交通驾驶员眼动特征及心电信号(ECG)的变化趋势,探究适宜的疲劳等级划分方法,构建有效的疲劳识别模型。方法招募6名志愿者进行城市轨道交通列车驾驶模拟实验,全程佩戴EQ02+型ECG数据采集套件和Dikablis眼镜式眼动仪,分析被试瞳孔面积变化,采用聚类分析法划分疲劳等级,分析被试SDNN、RMSSD、HF、HFnorm、LF/HF五项心电指标的变化,采用心电信号指标构建支持向量机(SVM)疲劳识别模型。结果 (1)被试瞳孔面积总体呈现波动下降的趋势,且与聚类分析法划分的疲劳等级具有一致性。(2)心电指标随实验时间变化明显,反映了交感神经和迷走神经的调节功能,可用于被试疲劳状态分析与识别。(3)对比三种不同核函数下模型的训练结果,高斯核函数识别效果最好,识别率达87.5%。结论眼动特征疲劳等级划分与心电指标SVM疲劳识别模型构建的结合,可为城市轨道交通驾驶员疲劳监测与预测的进一步研究提供依据。     

基于肝脏超声图像的纤维化量化分析

超声诊断除了受主观性和诊断经验等人为因素影响外,检测频率、机器参数的设置等成像条件的变化也会对图像产生很大的影响,因而使肝脏纤维化的量化分析受到很大限制.本文基于任意的肝脏超声图像提出了利用纹理边缘共生矩阵(TECM)进行纤维化量化分析的方法,首先由Canny算子提取纹理边缘,并计算其共生矩阵的熵作为分类的特征值.通过用leave-one-out最近邻法和Fisher线性分类器进行分类试验,证明其分类精度优于分形维(FD).而且,用Fisher线性分类器对TECM和差分盒计数法(DBC)组成的联合特征向量(J-DT)进行分类试验,当阈值为时,分类正确率(CCR)可以达到95.1%;取阈值为时,灵敏性可以达到100%.     

马来酸异丙酯-丙烯酸二元共聚物的合成研究

以马来酸异丙酯(MI)和丙烯酸(AA)为原料,采用活泼单体滴加的方式,通过水溶液自由基共聚合反应制备了MI/AA二元共聚物(PM-AA)。采用FT-IR对产物的结构进行了表征。通过正交实验和单因素实验对影响聚合反应的因素进行了考察和优化。在n(MI)∶n(AA)=1∶2,聚合温度85℃,AA滴加时间为1.5h,聚合时间2h,引发剂过硫酸铵用量为总单体质量的8%时,MI的转化率为76.17%。     

浅谈心脏磁场的检测及心脏源模型建模

比较心磁图(MCG)相对于传统心电图(ECG)的一些优势,介绍心脏磁场产生的机理,心脏磁场图的特点,SQUID检测的原理,并对心脏磁场各种检测方式的优势与不足做分析。同时,还介绍心磁图(MCG)分析,研究和建模的一些方法。最后对心脏磁场研究现状进行分析。     

基于YOLOv5s网络的垃圾分类和检测

目的 为了实现垃圾自动按类处理,通过研究基于视觉的垃圾检测与分类模型,实现对垃圾的自动识别和检测.方法 采用YOLOv5s网络作为垃圾检测与分类的模型,在自制垃圾分类数据集上对网络进行训练,利用训练好的YOLOv5s网络提取不同种类垃圾图像的特征和位置信息,实现垃圾的分类与检测.结果 在真实场景中进行了测试,基于YOLOv5s的垃圾分类检测模型可以有效识别6种不同形态的垃圾,检测mAP值为99.38％,测试精度为95.34％,目标检测速度达到6.67FPS.结论 实验结果表明,基于YOLOv5s网络的垃圾分类检测模型在不同光照、视角等条件下,检测准确率高,鲁棒性好、计算速度快.同时,有助于促进垃圾处理公司实现智能分拣,提高工作效率.     

A robust design method using variable transformation and Gauss–Hermite integration

Robust design seeks an optimal solution where the design objective is insensitive to the variations of input variables while the design feasibility under the variations is maintained. Accurate robustness assessment for both design objective and feasibility usually requires an intensive computational effort. In this paper, an accurate robustness assessment method with a moderate computational effort is proposed. The numerical Gauss–Hermite integration technique is employed to calculate the mean and standard deviation of the objective and constraint functions. To effectively use the Gauss–Hermite integration technique, a transformation from a general random variable into a normal variable is performed. The Gauss–Hermite integration and the transformation result in concise formulas and produce an accurate approximation to the mean and standard deviation. This approach is then incorporated into the framework of robust design optimization. The design of a two‐bar truss and an automobile torque arm is used to demonstrate the effectiveness of the proposed method. The results are compared with the commonly used Taylor expansion method and Monte Carlo simulation in terms of accuracy and efficiency. Copyright © 2005 John Wiley & Sons, Ltd.     

矩形高层建筑非高斯风压时程峰值因子计算方法

对一矩形高层建筑进行刚性模型风洞测压试验。首先基于风压时程的3阶矩和4阶矩,对不同风向下,矩形高层建筑表面风压脉动的高斯区与非高斯区进行划分。结果表明,迎风面的非高斯区主要位于迎风边缘及角部。在侧风面,长边迎风时,非高斯区主要集中在迎风前缘附近;短边迎风时,非高斯区位于靠近背风边缘的旋涡再附区。其次,基于Hermite级数法和改进Hermite级数法,通过参数分析和脉动风压谱拟合,提出非高斯风压时程的峰值因子简化计算方法--三参数Hermite级数法。最后,分别通过风压时程中的瞬时极值和风压时程的频数分布直方图验证了Hermite级数法、改进Hermite级数法和三参数Hermite级数法的有效性与准确性。结果表明,与风压时程中瞬时极值所对应的峰值因子相比,Hermite级数法的计算结果则偏于保守,三参数Hermite级数法、改进Hermite级数法的计算结果与其偏差率均小于10%。此外,三参数Hermite级数法和改进Hermite级数法所给出的概率密度曲线基本重合,且与实际风压时程频数分布直方图的贴合程度较优,能够较为准确地反映实际风压时程的概率密度分布。     

An automatic channel selection method based on the standard deviation of wavelet coefficients for motor imagery based brain–computer interfacing

The redundant data in multichannel electroencephalogram (EEG) signals significantly reduces the performance of brain–computer interface (BCI) systems. By removing redundant channels, a channel selection strategy increases the classification accuracy of BCI systems. In this work, a novel channel selection method (stdWC) based on the standard deviation of wavelet coefficients across channels is proposed to identify Motor Imagery (MI) based EEG signals. The wavelet coefficients are calculated by employing a Continuous Wavelet Transform (CWT) filter bank to decompose each trial from the EEG channel. The wavelet coefficient's standard deviation values are obtained across the channels, and these values are then sorted to determine the EEG channels with the highest standard deviation values. The channels with the largest wavelet coefficient divergence are chosen. MI trials are then spatially filtered with the Common Spatial Pattern (CSP), and CWT filter bank-based 2D images are generated from the spatially filtered trials. These images are then classified using a unique nine-layered convolutional neural network (CNN) model that combines two feature maps acquired with differing filter sizes. The proposed framework (stdWC-CSP-CNN) is evaluated using kappa score and classification accuracy on two publically accessible datasets (BCI Competition III dataset IVa and BCI Competition IV dataset 2a). The suggested framework achieved a mean test classification accuracy of 88.8% for dataset IVa from BCI Competition III and 75.03% for dataset 2a from BCI Competition IV, according to the results. The proposed channel selection method outperforms the other channel selection methods examined, according to the results. By rejecting redundant channels, the whole framework can improve the performance of MI-based BCIs.     

A Hermite reproducing kernel approximation for thin‐plate analysis with sub‐domain stabilized conforming integration

A Hermite reproducing kernel (RK) approximation and a sub‐domain stabilized conforming integration (SSCI) are proposed for solving thin‐plate problems in which second‐order differentiation is involved in the weak form. Although the standard RK approximation can be constructed with an arbitrary order of continuity, the proposed approximation based on both deflection and rotation variables is shown to be more effective in solving plate problems. By imposing the Kirchhoff mode reproducing conditions on deflectional and rotational degrees of freedom simultaneously, it is demonstrated that the minimum normalized support size (coverage) of kernel functions can be significantly reduced. With this proposed approximation, the Galerkin meshfree framework for thin plates is then formulated and the integration constraint for bending exactness is also derived. Subsequently, an SSCI method is developed to achieve the exact pure bending solution as well as to maintain spatial stability. Numerical examples demonstrate that the proposed formulation offers superior convergence rates, accuracy and efficiency, compared with those based on higher‐order Gauss quadrature rule. Copyright © 2007 John Wiley & Sons, Ltd.     

A Hermite reproducing kernel Galerkin meshfree approach for buckling analysis of thin plates

A Hermite reproducing kernel Galerkin meshfree approach is proposed for buckling analysis of thin plates. This approach employs the Hermite reproducing kernel meshfree approximation that incorporates both the deflectional and rotational nodal variables into the approximation of the plate deflection and the C ¹ continuous approximation requirement for the Galerkin analysis of thin plates can be easily achieved herein. The strain smoothing operation is consistently introduced to construct the smoothed rotation and curvature fields which appear in the weak form governing the thin plate buckling. The domain integration of the weak form is carried out by the method of sub-domain stabilized conforming integration with the smoothed measures of rotation and curvature, as leads to an efficient discrete meshfree formulation for the eigenvalue problem of thin plate buckling. A series of benchmark buckling problems are presented to assess the proposed algorithm and the results uniformly demonstrate the present approach is very effective and it performs superiorly compared to the conventional Galerkin meshfree formulations whose domain integration are performed by Gauss quadrature rules.     

MULTIVARIATE HERMITE APPROXIMATION FOR DESIGN OPTIMIZATION

An approximation based on multiple function and gradient information is developed using Hermite interpolation concepts. The goal is to build a high-quality approximation for complex and multidisciplinary design optimization problems employing analysis such as aeroservoelasticity, structural control, probability, etc. The proposed multidimensional approximation utilizes exact analyses data generated during the course of iterative optimization. The approximation possesses the property of reproducing the function and gradient information of known data points. The accuracy of the new approach is compared with linear, reciprocal and other standard approximations. Because the proposed algorithm uses more data points, its efficiency has to be compared in the context of iterative optimization.     

An improved method for R-peak detection by using Shannon energy envelope

R-peaks in electrocardiogram (ECG) play a vital role in diagnosis of heart rhythm irregularities and also estimating heart rate variability. However, almost all existing R-peak detectors suffer from the non-stationary of both QRS morphology and noise. To overcome these difficulties, we propose a four-stage improved method to detect R-peak using Shannon energy envelope. In the first stage, noise is suppressed and QRS complex is enhanced by using band pass filter, first order differentiation, and amplitude normalization. In the second stage, Shannon energy envelope is extracted. In the third stage, peak is estimated without considering any threshold amplitude. In the final stage, true R-peaks are detected. Our proposed R-peak detection method is validated using 48 first channel ECG records of the MIT-BIH arrhythmia database with the accuracy of 99.84%, sensitivity of 99.95% and positive predictability of 99.88%. Our proposed method outperforms other well-known methods in case of pathological ECG signals.     

非高斯随机过程的短期极值估计:复合Hermite模型

Hermite模型自20世纪80年代后期开始被广泛应用于非高斯随机过程的短期极值估计。当随机过程的非高斯性很强时,尤其是偏度很大时,常用的3阶Hermite模型不足以表征出极值分布的尾端特征。工程中,样本统计矩的不确定性使得更高阶的Hermite模型不宜使用。基于此,该文提出了同时基于中心矩与线性矩的复合Hermite模型,有效地将Hermite模型由3阶拓展到4阶。该文以对数正态模型作为非线性系统的研究对象,对比分析了在解析条件下和在使用蒙特卡洛模拟获得样本数据条件下,各类Hermite模型与传统的Gumbel法以及平均条件穿越率（ACER）法用于极值分析的表现。结果表明,对于大偏度强非高斯随机过程的极值预测,复合Hermite模型具有更好的精确度和鲁棒性。     

Meshless methods based on collocation with radial basis functions

Meshless methods based on collocation with radial basis functions (RBFs) are investigated in detail in this paper. Both globally supported and compactly supported radial basis functions are used with collocation to solve partial differential equations (PDEs). Using RBFs as a meshless collocation method to solve PDEs possesses some advantages. It is a truly mesh-free method, and is space dimension independent. Furthermore, in the context of scattered data interpolation it is known that some radial basis functions have spectral convergence orders. This study shows that the accuracy of derivatives of interpolating functions are usually very poor on boundary of domain when a direct collocation method is used, therefore it will result in significant error in solving a PDE with Neumann boundary conditions. Based on this fact, a Hermite type collocation method is proposed in this paper, in which both PDEs and prescribed traction boundary conditions are imposed on prescribed traction boundary. Numerical studies shows that the Hermite type collocation method improve the accuracy significantly. Received 31 January 2000