Automated delineation of non-small cell lung cancer: A step toward quantitative reasoning in medical decision science

Quantitative reasoning in medical decision science relies on the delineation of pathological objects. For example, evidence-based clinical decisions regarding lung diseases require the segmentation of nodules, tumors, or cancers. Non-small cell lung cancer (NSCLC) tends to be large sized, irregularly shaped, and grows against surrounding structures imposing challenges in the segmentation, even for expert clinicians. An automated delineation tool based on spatial analysis was developed and studied on 25 sets of computed tomography scans of NSCLC. Manual and automated delineations were compared, and the proposed method exhibited robustness in terms of the tumor size (5.32–18.24 mm), shape (spherical or irregular), contouring (lobulated, spiculated, or cavitated), localization (solitary, pleural, mediastinal, endobronchial, or tagging), and laterality (left or right lobe) with accuracy between 80% and 99%. Small discrepancies observed between the manual and automated delineations may arise from the variability in the practitioners' definitions of region of interest or imaging artifacts that reduced the tissue resolution.     

面向乳腺超声计算机辅助诊断的两阶段深度迁移学习

目的 为了提升基于单模态B型超声（B超）的乳腺癌计算机辅助诊断（computer-aided diagnosis,CAD）模型性能,提出一种基于两阶段深度迁移学习（two-stage deep transfer learning,TSDTL）的乳腺超声CAD算法,将超声弹性图像中的有效信息迁移至基于B超的乳腺癌CAD模型之中,进一步提升该CAD模型的性能。方法 在第1阶段的深度迁移学习中,提出将双模态超声图像重建任务作为一种自监督学习任务,训练一个关联多模态深度卷积神经网络模型,实现B超图像和超声弹性图像之间的信息交互迁移;在第2阶段的深度迁移学习中,基于隐式的特权信息学习（learning using privilaged information,LUPI）范式,进行基于双模态超声图像的乳腺肿瘤分类任务,通过标签信息引导下的分类进一步加强两个模态之间的特征融合与信息交互;采用单模态B超数据对所对应通道的分类网络进行微调,实现最终的乳腺癌B超图像分类模型。结果 实验在一个乳腺肿瘤双模超声数据集上进行算法性能验证。实验结果表明,通过迁移超声弹性图像的信息,TSDTL在基于B超的乳腺癌诊断任务中取得的平均分类准确率为87.84±2.08%、平均敏感度为88.89±3.70%、平均特异度为86.71±2.21%、平均约登指数为75.60±4.07%,优于直接基于单模态B超训练的分类模型以及多种典型迁移学习算法。结论 提出的TSDTL算法通过两阶段的深度迁移学习,将超声弹性图像的信息有效迁移至基于B超的乳腺癌CAD模型,提升了模型的诊断性能,具备潜在的应用可行性。     

钛合金植入物梯度孔结构设计及其力学性能

植入物多孔结构的设计研究多以规则孔结构为主,而少有针对梯度孔结构设计。本文提出平面center及空间sphere两种梯度圆孔设计方法,实现完成对某一平面center梯度孔的植入物结构设计,采用激光选区熔化技术(Selective Laser Melting, SLM)制备出孔隙率为75%的医用钛合金Ti6Al4V梯度孔与规则孔结构植入物样件,进行微观材料表征和力学性能测试,得到相关力学数据。结果表明：该种梯度孔结构的力学性能优于规则孔,在孔隙率为75%的条件下,梯度孔样件的平均弹性模量较规则孔高36.25%,平均抗压强度提高29.9%。     

Computer-Aided Targeting of the PI3K/Akt/mTOR Pathway: Toxicity Reduction and Therapeutic Opportunities

The PI3K/Akt/mTOR pathway plays an essential role in a wide range of biological functions, including metabolism, macromolecular synthesis, cell growth, proliferation and survival. Its versatility, however, makes it a conspicuous target of many pathogens; and the consequential deregulations of this pathway often lead to complications, such as tumorigenesis, type 2 diabetes and cardiovascular diseases. Molecular targeted therapy, aimed at modulating the deregulated pathway, holds great promise for controlling these diseases, though side effects may be inevitable, given the ubiquity of the pathway in cell functions. Here, we review a variety of factors found to modulate the PI3K/Akt/mTOR pathway, including gene mutations, certain metabolites, inflammatory factors, chemical toxicants, drugs found to rectify the pathway, as well as viruses that hijack the pathway for their own synthetic purposes. Furthermore, this evidence of PI3K/Akt/mTOR pathway alteration and related pathogenesis has inspired the exploration of computer-aided targeting of this pathway to optimize therapeutic strategies. Herein, we discuss several possible options, using computer-aided targeting, to reduce the toxicity of molecularly-targeted therapy, including mathematical modeling, to reveal system-level control mechanisms and to confer a low-dosage combination therapy, the potential of PP2A as a therapeutic target, the formulation of parameters to identify patients who would most benefit from specific targeted therapies and molecular dynamics simulations and docking studies to discover drugs that are isoform specific or mutation selective so as to avoid undesired broad inhibitions. We hope this review will stimulate novel ideas for pharmaceutical discovery and deepen our understanding of curability and toxicity by targeting the PI3K/Akt/mTOR pathway.     

A class of genetic algorithms for multiple-level warehouse layout problems

A new kind of warehouse layout problem, multiple-level warehouse layout problem (MLWLP), is investigated. Both horizontal and vertical travel costs need to be considered when making a layout. In the problem, unit travel costs are item-dependent and different items can be mixed in a cell. An IP model is proposed, which is shown to be NP-hard. An effective assignment method is presented and genetic algorithm heuristics developed. Extensive computational experiments are conducted to verify the effectiveness of the algorithms.     

A cutting sequence optimization algorithm to reduce the workpiece deformation in thin-wall machining

A thin-wall part of lower stiffness can be subject to significant deformation during its cutting process. This study proposes a cutting process optimization algorithm to reduce the workpiece deformation. First, the volume to be removed is divided into a set of blocks. The proposed algorithm starts from the finished workpiece shape, with all the blocks removed. The objective of the proposed algorithm is to find a sequence of adding the blocks, such that the workpiece deformation is always smaller than the given threshold value when the cutting forces is imposed at each step. The workpiece deformation at each step is simulated by using the FEM (finite element method) simulation. By inverting the sequence of adding the blocks, the optimized sequence to remove the blocks can be obtained. Additionally, the block size can be modified to reduce the axial depth of cut to further reduce the workpiece deformation, or to increase the radial depth of cut to enhance the efficiency. Experiments are conducted to confirm the effectiveness of the algorithm to reduce the maximum workpiece deformation during the entire cutting process.     

Exact 3D simulation of Scanning Electron Microscopy images of semiconductor devices in the presence of electric and magnetic fields

Simulations of Scanning Electron Microscopy images of semiconductor devices in the presence of electric fields are usually too simplistic, since they just rely on approximated solutions of the Poisson equation. In this paper, the 3D Poisson equation is solved in a TCAD environment, which accounts for realistic boundary conditions, as well as for complex physical effects like the formation of space charge regions in semiconductors and the polarization of dielectrics. The calculated solution is then passed to a Monte Carlo code that implements a new electron tracking engine optimized for speed, stability, and accuracy. After introducing the new tracking engine, three simulation examples are presented dealing with the presence of an extraction field, self-charging of the irradiated sample, and potential contrast in a biased silicon junction.     

地表水水源热泵系统计算机辅助优化设计研究综述

首先阐述了系统优化设计的类型和步骤,在此基础上介绍了国内外对水源热泵作为一个整体研究的情况和系统优化过程中的计算机应用情况;分析了系统各个部件的优化与建模现状;对目前常用的优化算法—遗传算法的概念及在热泵领域的应用进行了介绍。     

核磁共振波谱实验教学探讨

结合实例介绍了核磁共振谱实验在仪器分析教学中的应用,并在实验教学过程中运用计算机辅助教学对核磁共振氢谱进行解析。使理论教学和实验教学相结合,有利于提高学生分析解决问题的能力。