右美托咪啶在老年患者心脏手术中的临床应用

目的探讨右美托咪啶临床应用于老年患者心脏手术中的效果。方法采用随机对照试验,选取2015年7月-2017年6月该院行心脏手术的老年患者,将其分为实验组和对照组。实验组采用右美托咪啶、瑞芬太尼及丙泊酚联合行全身麻醉,对照组仅采用瑞芬太尼联合丙泊酚行全身麻醉;比较两组术中情况、手术效果及麻醉药物使用情况。结果实验组瑞芬太尼、丙泊酚的用量和输注速度均低于对照组(P<0.05);实验组SBP、DBP及MAP高于对照组,但HR低于对照组。结论相较于对照组,联合采用右美托咪啶行全身麻醉可使术中患者血流动力学表现更加稳定,可减少瑞芬太尼和丙泊酚用量,对患者术后康复具有积极意义。

基于多信息融合和卷积神经网络的行星齿轮箱故障诊断

基于机器学习的行星齿轮箱故障诊断方法依赖人工选择特征向量,而特征向量选择的优劣很大程度上决定了诊断方法的准确率。卷积神经网络(CNN)能自动提取特征,但用于行星齿轮箱故障诊断时难以通过单一振动信号对故障做出精确诊断。针对上述问题,提出了一种基于多信息融合和CNN的行星齿轮箱故障诊断方法。对行星齿轮箱的三向(水平径向、垂直径向与轴向)振动信号和声音信号进行数据层融合,将一维的振动信号和声音信号通过并联方式整合为一个二维信号;将二维信号作为CNN的输入,利用多个卷积层和最大池化层进行深度特征提取和信息过滤,最终通过Softmax分类器实现故障分类。搭建了行星齿轮箱故障诊断实验台,采集不同转速和负载工况下行星齿轮箱正常和故障状态的振动信号和声音信号,并输入CNN中进行训练和验证。在相同条件下选取水平径向振动信号、垂直径向振动信号、轴向振动信号、声音信号4种单源信息分别与CNN相结合的方法进行对比,以验证基于多信息融合和CNN的行星齿轮箱故障诊断方法的优越性,实验结果表明:轴向振动信号+CNN和声音信号+CNN 2种方法的故障识别准确率分别为74.07%和75.13%;水平径向振动信号+CNN和垂直径向振动信号+CNN 2种方法的故障识别准确率分别为89.70%和87.09%;基于多信息融合和CNN方法的收敛速度最快,故障识别准确率最高,为93.33%。

CD4+ Tcell and tuberculosis results of AIDS patients with different types of tuberculosis

Objective: To investigate the correlation of CD4+T cell and tuberculosis test in AIDS patients with different types of tuberculosis. Methods: A total of 127 cases of AIDS complicated with tuberculosis were selected from our hospital between June 2014 and June 2017 as observation group;50 cases with non-tuberculosis respiratory system who were admitted to our hospital during the same period were selected as control group. The positive rate of TB was examined by T.spotTB test and TST test, and the level of CD4+T lymphocyte was measured by flow cytometry. Results: The positive rate of the observed group After T.spotTB test and TST test was higher than the control group;observation group CD4+T lymphocyte level of 0-100/mm3 was significantly higher than that of the control group;different types of tuberculosis with AIDS showed no significant difference between the T.spotTB test and the TST test in terms of positive rate;There was also no significant difference between the different types of tuberculosis with HIV CD4+T lymphocyte level;the levels of 100-200/mm3 and >200/mm3 CD4+T lymphocytes and T.spotTB test positive and TST positive test were negatively correlated, while the levels of CD4+T lymphocyte 0-100/mm3 and positive T.spotTB test and TST test were positively correlated. Conclusions: the lower level of CD4+T cells in AIDS patients with different types of tuberculosis is positively related to the positive T.spotTB test and TST test.

基于薄板样条插值的弯曲笔触神经绘画与风格化方法

近年来,图像生成技术取得了令人瞩目的发展,目前的图像生成方法大多以像素填充的方式生成图像,缺乏艺术家逐笔绘制的过程,使其在局部笔触细节与纹理上与真实艺术作品有所不同.神经绘画致力于模仿人类艺术家以画笔为单位,逐笔绘制的作画过程.现有的神经绘画方法大多使用贝塞尔曲线或者笔触模板进行仿射变换来模拟真实笔触.然而,贝塞尔曲线纹理的缺乏以及仿射变换的线性性质,导致生成的笔触在纹理或者形状上存在较大的限制.为了更好地模拟真实笔触的纹理与形状,本文提出了新的基于薄板样条插值的弯曲笔触参数模型,通过对真实笔触模板先后进行弯曲与仿射变换,可以生成更加真实、多样的笔触图像.此外,本文提出了层次化的笔触优化方法,将整幅图像分解为由大到小的多个笔触,能够有效提升模型对图像整体架构与局部细节的绘画能力.最后,本文将提出的方法拓展至风格迁移中,实现了较好的风格迁移效果.定性与定量的实验表明,本文所提出的新的笔触模型与优化方法在神经绘画及风格化任务中都超越了已有的最佳模型.

Biochar addition to tea garden soils:effects on tea fluoride uptake and accumulation

Long-term consumption of tea with high fluoride(F)content has a potential threat to human health.The application of different amounts of biochar to reduce F accumulation in tea leaves has been little studied.In this study,a pot experiment was conducted to investigate the effect of biochar amounts(0,0.5%,2.5%,5.0%,8.0%,and 10.0%,w/w)on tea F content during the tea plant growth.Changes in tea quality,soil F fraction,and soil properties caused by biochar and the relationship with tea F accumulation were also considered.The results showed that the application of biochar amendment significantly reduced water-soluble F contents in tea leaves compared to CK(without biochar),especially in the 8.0%treatment(72.55%).Overall,biochar contributed to improving tea polyphenols and caffeine,but had no significant impact on free amino acids and water leachate.Compared with CK,5.0-10.0%biochar significantly increased soil water-soluble F content due to the substitution of F−with OH−under high pH.Additionally,biochar applied to tea garden soil was effective in decreasing the soil exchangeable aluminum(Ex-Al)content(46.37-91.90%)and increasing the soil exchangeable calcium(Ca^(2+))content(12.02-129.74%)compared to CK,and correlation analysis showed that this may help reduce F enrichment of tea leaves.In general,the application of 5.0-8.0%biochar can be suggested as an optimal application dose to decrease tea F contents while simultaneously improving tea quality.

Implicit privacy preservation:a framework based on data generation

This paper addresses a special and imperceptible class of privacy,called implicit privacy.In contrast to traditional(explicit)privacy,implicit privacy has two essential prop-erties:(1)It is not initially defined as a privacy attribute;(2)it is strongly associated with privacy attributes.In other words,attackers could utilize it to infer privacy attributes with a certain probability,indirectly resulting in the disclosure of private information.To deal with the implicit privacy disclosure problem,we give a measurable definition of implicit privacy,and propose an ex-ante implicit privacy-preserving framework based on data generation,called IMPOSTER.The framework consists of an implicit privacy detection module and an implicit privacy protection module.The former uses normalized mutual information to detect implicit privacy attributes that are strongly related to traditional privacy attributes.Based on the idea of data generation,the latter equips the Generative Adversarial Network(GAN)framework with an additional discriminator,which is used to eliminate the association between traditional privacy attributes and implicit ones.We elaborate a theoretical analysis for the convergence of the framework.Experiments demonstrate that with the learned gen-erator,IMPOSTER can alleviate the disclosure of implicit privacy while maintaining good data utility.

Analysis of the effects of the heat of sorption on the process of heat transfer in moisture exchange across a membrane

A theoretical study has been conducted to investigate the effects of mass transfer on heat transfer in moisture exchange across a membrane and a mathematical model describing the heat transfer process with consideration of the heat of sorption was established. A dimensionless variable, Ψ =JLλ /δ (T 10−T 20), which controls the effect of the heat of sorption on the heat transfer in membrane process, was obtained through theoretical analysis, and the effects of Ψ on the heat transfer process were analyzed. Results showed that in the case that the temperature gradient and mass transfer are in the same direction, the effective heat flux changes the direction at Ψ=1. For Ψ【1, the heat transfers from high to low temperature sides through the membrane, and the total thermal resistance increases with increasing the moisture flux across membrane or reducing the temperature difference between the bulk flows on the two sides of membrane. For Ψ】1, the overall effect of the heat and mass transfer is that the effective heat flux points from low to high temperature sides and the total thermal resistance decreases with increasing the mass flux or reducing the temperature difference. In the case that the temperature gradient and mass transfer are in the opposite directions, the existence of the heat of sorption acts to enhance the heat transfer from high to low temperature sides, causing a reduced total thermal resistance, and the greater the mass flux or the smaller the temperature difference, the smaller the total thermal resistance.

An integrative model and analysis of cell cycle in fission yeast

According to the recent investigation on cell cycle of fission yeast, a mathematical dynamic model is for-mulated. Four cyclins, e.g. Puc1, Cig1, Cig2 and Cdc13, are investigated here. The interacting networks between the cy-clins and the process of cell cycle are mathematically de-scribed. The functions of these cyclins are particularly ana-lyzed. Comparison among different mutants indicates that the cyclins play an important role in cell cycle.

Implicit privacy preservation:a framework based on data generation

This paper addresses a special and imperceptible class of privacy,called implicit privacy.In contrast to traditional(explicit)privacy,implicit privacy has two essential properties:(1)It is not initially de ned as a privacy attribute;(2)it is strongly associated with privacy attributes.In other words,attackers could utilize it to infer privacy attributes with a certain probability,indirectly resulting in the disclosure of private information.To deal with the implicit privacy disclosure problem,we give a measurable de nition of implicit privacy,and propose an ex-ante implicit privacy-preserving framework based on data generation,called IMPOSTER.The framework consists of an implicit privacy detection module and an implicit privacy protection module.The former uses normalized mutual information to detect implicit privacy attributes that are strongly related to traditional privacy attributes.Based on the idea of data generation,the latter equips the Generative Adversarial Network(GAN)framework with an additional discriminator,which is used to eliminate the association between traditional privacy attributes and implicit ones.We elaborate a theoretical analysis for the convergence of the framework.Experiments demonstrate that with the learned generator,IMPOSTER can alleviate the disclosure of implicit privacy while maintaining good data utility.