Aptamer biosensor for label-free impedance spectroscopy detection of thrombin based on gold nanoparticles

This paper presents a simple electrochemical impedance spectroscopy (EIS) aptasensor based on an anti-thrombin-aptamer as a molecular recognition element. Improvement in sensitivity was achieved by utilizing gold nanoparticles (AuNPs), which were self-assembled on the surface of a bare electrode by using 1,6-Hexanedithiol as a medium. To quantify the amount of thrombin, changes in the interfacial electron transfer resistance (R_et) of the aptasensor were monitored using the redox couple of an [Fe(CN)₆]^3−/4− probe. The plot of (Reti−Ret₀)/Ret₀ against the logarithm of thrombin concentration is linear with over the range from 0.1 nM to 30 nM with a detection limit of 0.013 nM. Meanwhile, the packing density of aptamers was determined by cyclic voltammetric (CV) studies of redox cations (e.g., [Ru(NH₃)₆]³⁺) which were electrostatically bound to the DNA phosphate backbones. The results indicate that the total amount of aptamer probes immobilized on the gold nanoparticle surface is sixfold higher than that on the bare electrode. The aptasensor also showed good selectivity for thrombin without being affected by the presence of other proteins.     

A distributed Bragg reflector porous silicon layer for optical interferometric sensing of organic vapor

In this study, we successfully demonstrated the rapid, sensitive, and reversible sensing of organic vapor using a distributed Bragg reflector (DBR) porous silicon (PS) layer. We fabricated the DBR PS layer on a p⁺-type silicon substrate and investigated its reflectance spectra before, during, and after exposure to the different concentrations of various organic vapors. When the DBR PS layer sample was exposed to methanol, acetone, ethanol, and isopropanol vapors, the maximum reflectance peak promptly shifted toward longer wavelengths by about 4.5, 23.2, 26.0, and 38.2 nm, respectively. We determined that the red-shift in the reflectance spectrum could be attributed to the changes in the refractive index induced by the capillary condensation of the organic vapor within the pores of the DBR PS layer. The DBR PS layer showed excellent sensing ability under the different concentrations and types of organic vapors. In addition, a slight hysteresis of the red-shift was observed during repeated exposure to organic vapors at different concentrations. After removing the organic vapors, the reflectance spectrum promptly returned to its original state.     

电阻抗扫描成像计算机辅助诊断方法研究

为了提高电阻抗扫描成像EISI(Electrical Impedance Scanning Imaging)在乳腺疾病检查方面的准确性,将独立分量方法(ICA)应用于EISI数据的特征提取,并结合支持向量机(SVM)进行乳腺疾病计算机辅助诊断(CAD)。实验结果表明:ICA方法所提取的特征对于乳腺疾病分类是有效的。对于良性肿瘤和恶性肿瘤的鉴别,提出的CAD方法灵敏度为74.2%,特异度为82.8%,准确度为80%。结合所提出的CAD方法和传统的EISI检查方法,可以提高EISI在乳腺疾病检查方面的可靠性,降低假阳性率,避免对良性肿瘤患者进行活检。     

改进的证据理论在合作频谱感知中的应用

为了能够在认知无线电网络中有效、准确地检测出授权用户,提出了基于改进的证据理论的合作频谱感知方案设计方法.根据各个次用户对授权用户进行多次本地检测得出的时域DS融合结果,设计并改进融合顺序和融合结构,将多个时域DS融合结果送到融合中心进行空域DS融合,得到最终判决.在此基础上,给出了融合顺序、融合结构的改变对算法性能的...     

采用商业卫星构建虚拟星座

提出了虚拟星座的概念,从商业卫星资源优势、卫星参数差异和传感器特征差异3个方面分析了构建虚拟星座的可行性。以LANDSAT-7卫星和SPOT5卫星为试验对象,验证了构建虚拟星座的可操作性。     

基于NHibernate的数据持久化方案

分析目前广泛使用的对象关系映射（ORM）组件——NHibernate,针对NHibernate的不足提出改进的ORM模型,根据该模型设计并实现基于.NET平台的轻量级ORM组件——MYPL,利用该组件无须编写任何映射文件即可实现对象关系映射等数据持久化操作,并能降低业务层和数据层之间的耦合,提高系统的扩展性、可维护性和应用系统的开发效率。     

基于近红外与高光谱技术的小麦种子多指标检测方法

采集47份小麦样本的高光谱图像,提取感兴趣区内的平均光谱,结合反映小麦品质的水分、蛋白质和湿面筋三个指标基础数据,通过OPUS软件搜寻最佳的光谱预处理和波段组合,分别建立了三个指标的偏最小二乘法(PLS)定量分析模型.同时与采用近红外(.NIR)光谱技术建立的三个指标的近红外模型进行比较,发现高光谱模型的各性能指标均明显优于近红外模型.结果表明:当被测样品为颗粒状且内部化学成分分布不均匀时,近红外模型的准确性和稳定性会受其测量条件的限制,而高光谱采样面积大,获取信息更全面,展现出强大的分析检测潜质,为小麦品质评价提供了新方法.     

强干扰环境下高炉雷达信号机器学习算法

高炉料面属矿物-煤气-焦炭多元高温固体/熔体混杂共存的粗糙表面,其电磁反射特征包含非均匀和非平稳的气固混合介质的表层电磁回波信息、布料溜槽引起的周期性遮蔽效应、十字测温等装置引起的固定干扰,以及电磁辐射等环境因素引发的随机噪声.本文研究了复杂环境下调频连续波(FMCW)提取的料面信号,采用瞬时频率分析的希尔伯特-黄变换(HHT)方法代替传统快速傅里叶变换(FFT)方法.结合经验模态分解,将原始非平稳信号分解为若干个平稳的内在模式函数;并按照基于先验知识的决策树算法分类与学习,获得各类分量权值并加权突出真实物料的电磁信号;通过Hilbert变换得到原始信号的时频域特征,可以揭示流态化料面包含的丰富冶炼信息.同时该算法也有助于提高料面成像的帧准确率和稳定性,为钢铁行业节能减排提供可靠的数据支撑.     

基于支持向量机的无创血糖光谱算法

为了实现无创血糖浓度检测,提出了基于支持向量机回归模型的无创血糖光谱算法. 该算法使用光电容积脉搏波（PPG）设备对志愿者指端红光、红外光交替采样得到PPG信号,然后通过微创血糖仪测得血糖浓度. 对采集到的PPG信号进行处理提取特征组成特征矩阵,分别运用不同机器学习模型对特征矩阵和实时血糖浓度进行回归训练,得到特征矩阵与血糖浓度间的关系,并对训练得到的函数关系进行验证,选取出高斯核支持向量机模型为最佳训练模型. 实验证明,与偏最小二乘回归进行对比,本文提出的运用核函数为高斯核的支持向量机算法的预测准确度能提升10%～15%,预测的高低血糖正确率达到98%.