基于自监督学习的病理图像层次分割

在苏木精-伊红(HE)染色病理图像中，细胞染色分布的不均匀和各类组织形态的多样性给病理图像的自动分割带来极大挑战。为解决该问题，提出了一种基于自监督学习的病理图像三步层次分割方法，对病理图像中各类组织进行由粗略到精细的全自动逐层分割。首先，根据互信息的计算结果在RGB色彩空间中进行特征选择；其次，采用K-means聚类将图像初步分割为各类组织结构的色彩稳定区域与模糊区域；然后，以色彩稳定区域为训练集采用朴素贝叶斯分类对模糊区域进行进一步分割，得到完整的细胞核、细胞质和胞外间隙这三类组织结构；最后，对细胞核部分进行结合形状和色彩强度的混合分水岭分割得到细胞核间的精确边界，进而量化计算细胞核个数、核占比、核质比等指标。对脑膜瘤HE染色病理图像的分割实验结果表明，所提方法对于染色和细胞形态差异保持较高的鲁棒性，各类组织区域分割误差在5%以内，在细胞核分割精度的对比实验中平均正确率在96%以上，满足临床自动图像分析的要求，其量化结果可以为定量病理分析提供依据。     

Inverse identification of constitutive parameters from heat source fields: A local approach applied to hyperelasticity

In this paper, a new inverse identification method of constitutive parameters is developed from full kinematic and thermal field measurements. It consists in reconstructing the heat source field from two different approaches by using the heat diffusion equation. The first one requires the temperature field measurement and the value of the thermophysical parameters. The second one is based on the kinematic field measurement and the choice of a thermo-hyperelastic model that contains the parameters to be identified. The identification is carried out at the local scale, ie, at any point of the heat source field, without using the boundary conditions. In the present work, the method is applied to the challenging case of hyperelasticity from a heterogeneous test. Due to large deformations undergone by the rubber specimen tested, a motion compensation technique is developed to plot the kinematic and the thermal fields at the same points before reconstructing the heterogeneous heat source field. In the present case, the constitutive parameter of the Neo-Hookean model has been identified, and its distribution has been characterized with respect to the strain state at the surface of a cross-shaped specimen.     

Effects of food container depth on the quality and yield of superchilled and iced Atlantic salmon

Insulated food containers of different sizes are commonly used to preserve the quality of perishable fresh fish products during transport and storage. The aim of the study was to explore the effects of packaging solutions with various depths on the quality of iced and superchilled farmed Atlantic salmon (Salmo salar). Insulated containers of different depths (32, 42, and 60 cm) and 17-cm deep expanded polystyrene (EPS) boxes were used to simulate transport and storage of the fish. The fish quality was evaluated after 4, 10, and 14 days of storage, where measurements on drip loss, texture, cooking yield, sensory evaluation, water and lipid distribution, and characteristics analysis in the fish muscle by nuclear magnetic resonance (NMR) spectroscopy were performed. Increased container depth from 32 to 42 cm increased the drip loss of superchilled salmon from 2.4% to 3.1% after 10 days of storage at −1°C. Iced storage of salmon for 10 days in EPS boxes resulted in less drip loss (0.5%) compared with superchilled salmon (1.5%) stored in EPS. The NMR analysis, sensory evaluation, texture analysis, and cooking yield did not reveal any major differences between salmon stored in containers of different depths. Iced storage increased pressure marks as compared with non-iced, superchilled storage in insulated containers. The results indicate that transport of whole salmon in reusable insulated plastic containers with a depth of 60 cm is possible without major quality changes observed for up to 10 days.     

基于声源定位的高校在线巡课系统设计与研究

随着网络建设以及信息化教学方法在高校教学过程中的应用普及,越来越多的高校使用在线巡课系统对教师的教学过程进行跟踪和管理,以便发现课堂教学中的亮点、问题和不足。文章提出的在线巡课系统,基于声源定位的技术,对传统的在线巡课系统做出了改进,解决了已有巡课系统中“只闻其声,不见其人”的问题,能够更直观地跟踪到教师的教学过程,包括教学行为以及师生互动过程,有效提升教务人员巡课效果和体验感受。     

A self-adaptive synthetic over-sampling technique for imbalanced classification

Traditionally, in supervised machine learning, (a significant) part of the available data (usually 50%-80%) is used for training and the rest—for validation. In many problems, however, the data are highly imbalanced in regard to different classes or does not have good coverage of the feasible data space which, in turn, creates problems in validation and usage phase. In this paper, we propose a technique for synthesizing feasible and likely data to help balance the classes as well as to boost the performance in terms of confusion matrix as well as overall. The idea, in a nutshell, is to synthesize data samples in close vicinity to the actual data samples specifically for the less represented (minority) classes. This has also implications to the so-called fairness of machine learning. In this paper, we propose a specific method for synthesizing data in a way to balance the classes and boost the performance, especially of the minority classes. It is generic and can be applied to different base algorithms, for example, support vector machines, k-nearest neighbour classifiers deep neural, rule-based classifiers, decision trees, and so forth. The results demonstrated that (a) a significantly more balanced (and fair) classification results can be achieved and (b) that the overall performance as well as the performance per class measured by confusion matrix can be boosted. In addition, this approach can be very valuable for the cases when the number of actual available labelled data is small which itself is one of the problems of the contemporary machine learning.     

Parameter tuning Naïve Bayes for automatic patent classification

I present an analysis of feature selection for automatic patent categorization. For a corpus of 7,309 patent applications from the World Patent Information (WPI) Test Collection (Lupu, 2019), I assign International Patent Classification (IPC) section codes using a modified Naïve Bayes classifier. I compare precision, recall, and f-measure for a variety of meta-parameter settings including data smoothing and acceptance threshold. Finally, I apply the optimized model to IPC class and group codes and compare the results of patent categorization to academic literature.     

Study of equivalent efficiency calibration method for radioactive gas source measured on HPGe detectors

ABSTRACT

In view of the complexity of current detection efficiency calibration of radioactive gas sources, a method using solid planar sources to be equivalent to gas sources was studied. For the 50 mL gas source box, an optimal equivalent scheme was selected by Monte Carlo Simulations. Then, the full-energy-peak efficiency curve of gas sources at the measurement position of 25 cm, with source-to-detector distance of 25 cm, was fitted by measuring solid planar sources with known activity. To verify the accuracy of the efficiency curve, ⁴¹Ar, ¹³³Xe and ⁸⁷Kr gases were produced and determined by length-compensated method. Then, their full-energy-peak efficiencies at 25 cm position away from the detector were directly calibrated. The percentage efficiency deviations between interpolation from the efficiency curve and direct calibration are all less than 2.5%, which proves the accuracy of the equivalent method. This calibration method is a general one and can be also used for some other radioactive sample measurements, such as non-destructive analysis of gaseous fission product samples with a suitable source-to-detector distance.     

Water leakage due to the welding defect and improvement to reach 1-MW beam operation in the mercury target of J-PARC

ABSTRACT

Neutron flux per pulse reached world record at neutron source in Japan Proton Accelerator Research Complex (J-PARC). In the J-PARC, mercury target system is used as a spallation neutron source. A target vessel has a multi-walled protection system that comprises a mercury vessel enclosed with a double-walled water shroud. This is to prevent the leakage of the mercury outside the target vessel. The multi-walled structure needed to be complicated with a lot of welding lines. However, during the operation, we faced an unscheduled shutdown due to water leakage to the intermediate layer between the mercury vessel and water shroud. An investigation on the cause of the leakage was carried out. It is deduced that the leakage path was formed due to the crack propagation from welding defects that are caused by the complicated multi-walled structure. The crack propagation is attributed to the repeated stress by pressure waves generated in the mercury target. Based on the investigation results, the design was improved to remove the welding line on the complicated structure and to realize the stable operation with 1 MW proton beam, which is the final design value of the neutron source in J-PARC.     

COB LED自动在线点胶测试补粉机的研制

在COB(Chip On Board)LED光组件制造领域,点胶、测试和补粉关键工艺基本以手工或半自动为主。为了保证COB LED产品质量一致性、提高生产效率,研制了一种COB自动在线点胶测试及补粉一体机,实现点胶、测试、补粉生产工序的自动化生产。阐述了其工艺流程及设计总体方案,提出了荧光粉量修补方法,修正初次点胶量,实现动态调节,减少人工干预。经实验验证,该一体机可以提高点胶工艺的自动化程度,具有运行速度快、可靠性好、性能稳定等特点。     

油藏开发地质勘探的常见问题分析

就油藏开发地质勘探的常见问题进行分析,从油藏开发的分类、油藏开发程序以及油藏储量计算等方面综合考虑影响油藏开发的因素,为现实油藏开发工程提供理论依据,提升油藏开发技术水准。