Integrating expert knowledge with data in Bayesian networks: Preserving data-driven expectations when the expert variables remain unobserved

When developing a causal probabilistic model, i.e. a Bayesian network (BN), it is common to incorporate expert knowledge of factors that are important for decision analysis but where historical data are unavailable or difficult to obtain. This paper focuses on the problem whereby the distribution of some continuous variable in a BN is known from data, but where we wish to explicitly model the impact of some additional expert variable (for which there is expert judgment but no data). Because the statistical outcomes are already influenced by the causes an expert might identify as variables missing from the dataset, the incentive here is to add the expert factor to the model in such a way that the distribution of the data variable is preserved when the expert factor remains unobserved. We provide a method for eliciting expert judgment that ensures the expected values of a data variable are preserved under all the known conditions. We show that it is generally neither possible, nor realistic, to preserve the variance of the data variable, but we provide a method towards determining the accuracy of expertise in terms of the extent to which the variability of the revised empirical distribution is minimised. We also describe how to incorporate the assessment of extremely rare or previously unobserved events.     

医学图像分析深度学习方法研究与挑战

深度学习（Deep learning,DL）,特别是深度卷积神经网络（Convolutional neural networks,CNNs）,能够从医学图像大数据中自动学习提取隐含的疾病诊断特征,近几年已迅速成为医学图像分析研究热点.本文首先简述医学图像分析特点;其次,论述深度学习基本原理,总结深度CNNs在医学图像分析中的分类、分割框架;然后,分别论述深度学习在医学图像分类、检测、分割等各应用领域的国内外研究现状;最后,探讨归纳医学图像分析深度学习方法挑战及其主要应对策略和开放的研究方向.     

一种新的基于体元剖分的软组织切割算法

切割是虚拟手术系统中需要表现的真实动作之一,真实感和实时性是衡量切割效果的关键.因此有必要研究一种新的基于体元剖分的软组织切割算法.该算法不同于以往的体元剖分算法,它将整个剖分过程分两步来完成：首先通过退化处理过程将不完全切割的四面体单元分解成多个完全切割的四面体单元;然后将切割情况进行分类,根据不同的切割情况分别进行相应的基元分解.实验表明,研究的新算法用于四面体单元的有限元模型,具有良好的切割处理效果.     

鱼香肉丝在烹饪前后营养成分的评价

为系统评价中国传统菜肴鱼香肉丝的营养品质以及烹饪前后的营养成分变化,将用于制作鱼香肉丝的原材料平均分成2组,每组3份,分别是生原料组和成品组。对其22种营养素指标进行检测,并分别对常规营养素、氨基酸、脂肪酸、矿质元素以及维生素进行营养评价。结果表明,在烹饪后鱼香肉丝的蛋白质显著升高(P<0.05),脂肪、磷、和铁的含量极显著升高(P<0.01),但氨基酸组成和脂肪酸总量无显著变化,维生素B₁显著降低(P<0.01)。蛋白质的氨基酸组成合理,必需氨基酸指数达到了85.91%;且多不饱和脂肪酸含量丰富,占总脂肪酸的58.36%。此外,矿质元素的营养质量指数均大于1(除钙以外),基本满足人体需求,但鱼香肉丝的维生素A和维生D含量较为缺乏,B族维生素(B₁、B₃、B₁₂)含量也低于中国居民膳食营养素摄入量的推荐值。综上,鱼香肉丝是一道营养丰富、均衡的菜肴,它具有优质的蛋白质、脂肪酸和丰富的矿质元素。烹饪会造成其维生素B₁的损失,而大大增加铁元素的含量,并且蛋白质和脂肪的含量会升高。本研究为膳食记录提供了更为准确的营养数据。     

AES加密算法分析及其在信息安全中的应用

随着现代密码分析水平、芯片处理能力的不断提高,高级加密标准AES算法将在各行业广泛应用。该文章介绍了AES加密算法的发展以及实现过程,并探讨其在信息安全方面的应用。     

Improving the incorporation of indium component for InGaN-based green LED through inserting photonic crystalline in the GaN layer

We report on the effect of inserted photonic crystalline (Ph-C) in the GaN epitaxial layer on the incorporation of the indium component for the InGaN-based green LED. The adoption of Ph-C in the GaN layer shifted the Raman peak value of E₂ mode of GaN to lower frequency and resulted in a tensive stress relief. The stress relief can be attributed to strained lattices restoring in the matrix of Ph-C and the GaN pseudo-epitaxy over the air-void of the Ph-C. Moreover, the HRXRD rocking curves and AFM results show that the insertion of Ph-C also improves the crystal quality. With the inserted Ph-C, the indium component in the multiple quantum wells of the green LED (Ph-C LED) was enhanced. This resulted in a 6-nm red-shift of the peak wavelength. Furthermore, the LOP of the Ph-C LED was enhanced by 10.65% under an injection current of 20 mA.     

Silk fibroin/chitosan scaffold: preparation,characterization, and culture with HepG2 cell

Tissue engineering requires the development of three-dimensional water-stable scaffolds. In this study, silk fibroin/chitosan (SFCS) scaffold was successfully prepared by freeze-drying method. The scaffold is water-stable, only swelling to a limited extent depending on its composition. Fourier Transform Infrared (FTIR) spectra and X-Ray diffraction curves confirmed the different structure of SFCS scaffolds from both chitosan and silk fibroin. The homogeneous porous structure, together with nano-scale compatibility of the two naturally derived polymers, gives rise to the controllable mechanical properties of SFCS scaffolds. By varying the composition, both the compressive modulus and compressive strength of SFCS scaffolds can be controlled. The porosity of SFCS scaffolds is above 95% when the total concentration of silk fibroin and chitosan is below 6 wt%. The pore sizes of the SFCS scaffolds range from 100 μm to 150 μm, which can be regulated by changing the total concentration. MTT assay showed that SFCS scaffolds can promote the proliferation of HepG2 cells (human hepatoma cell line) significantly. All these results make SFCS scaffold a suitable candidate for tissue engineering.