Search

Introduction: Recently, deep learning based methods have become an important approach to the accurate analysis of medical images. Methods and Materials: This paper provides a comprehensive survey of the most important deep learning based methods that have been developed for medical image processing. A number of important contributions made in the last five years are summarized and surveyed. Results: Specifically, deep learning-based algorithms developed for image segmentation, image classification, registration, object detection and other important problems are reviewed. Conclusion: In addition, an overview of challenges that currently exist in the field and potential directions for future research is provided at the end of the survey.

Effects of glycerin and silica sol on the structure and performance of porous Si3N4 ceramics that were prepared via freeze-drying were studied. Adding glycerol affected growth process of ice crystals, changed pore structural from bimodal peaks to triple peaks distribution, and formed regular honeycomb structure. This improved mechanical properties and electromagnetic wave transmission properties of Si3N4 ceramics. Therefore, adding glycerol changed mesoscopic morphology of the material. For porous Si3N4 ceramics with 72.5% porosity, flexural strength was 38 MPa, compressive strength was 42.6 MPa, and dielectric constant was 1.75 in the frequency range of 12.4–18 Hz as the amount of added glycerol increasing to 5 wt%. By reducing the sintering eutectic point of the system, the addition of silica sol promoted sintering shrinkage and formed new phases. Thus, pore size distribution was changed from bimodal to unimodal. Moreover, dielectric constant of samples increased from 2.25 to 3.5. Finally, structure and performance of porous Si3N4 ceramics were controlled via addition of glycerol and silica sol.

Background: Image segmentation is an important problem in both image processing and computer vision. Given an image, the goal of image segmentation is to label each pixel in the image such that the pixels with the same label collectively represent an object. Materials and Methods: Due to the inherent complexity and noise that may exist in images, developing an algorithm that can generate excellent segmentation results for an arbitrary image is still a challenging problem. In this paper, a new adaptive Hidden Markov Model is developed to describe the spatial and semantic relationships among pixels in an image. Based on this statistical model, image segmentation can be efficiently performed with an adaptive Viterbi algorithm in linear time. Results: The algorithm is unsupervised and does not require being used along with any other approach in image segmentation. Testing results on synthetic and real images show that this algorithm is able to achieve excellent segmentation results in both ordinary images and medical images. Conclusion: An implementation of this algorithm in MATLAB is freely available upon request.

Heavy metals accumulate in high water table coal mining subsidence ponds, resulting in heavy metal enrichment and destruction of the ecological environment. In this study, subsidence ponds with different resource reutilization methods were used as study subjects, and non-remediated subsidence ponds were collectively used as the control region to analyze the heavy metal distributions in water bodies, sediment, and vegetation. The results revealed the arsenic content in the water bodies slightly exceeded Class III of China&rsquo, s Environmental Quality Standards for Surface Water. The lead content in water inlet vegetation of the control region and the Anguo wetland severely exceeded limits. Pearson&rsquo, s correlation, PCA, and HCA analysis results indicated that the heavy metals at the study site could be divided into two categories: Category 1 is the most prevalent in aquaculture pond B and mainly originate from aquaculture. Category 2 predominates in control region D and mainly originates from atmospheric deposition, coal mining, and leaching. In general, the degree of heavy metal contamination in the Anguo wetland, aquaculture pond, and fishery&ndash, solar hybrid project regions is lower than that in the control region. Therefore, these models should be considered during resource reutilization of subsidence ponds based on the actual conditions.

Effects of Ni and B additions on the microstructure and growth behavior of the intermetallic compound (IMC) of Sn–1.0Ag–0.5Cu alloys (SAC105) were investigated in this study. Results show that microadditions of Ni and B result in volume fraction of primary Sn increasing and the grain size decreasing observably. It is found that a large number of fine reinforcement particles with network-like shape are found in the solder, and the thickness of interfacial IMC layer in the solder joint is grew less than that of SAC105 with longer aging time. Shear test results reveal that as-soldered solder joints of microalloyed SAC105 have better shear strength than that of SAC105 solder alloy.

Four rock assemblages in correspondence with tow different tectionic settings have been recognized in the NEE-SWW extending HP-UHP metarmorphic belt in southwestern Tianshan, northwest China, Eclogite assemblage ECI is geochemically akin to alkaline within-plate oceanic island basalt (OIB). EC2 shows affinity to enriched mid-oceanic ridge basalt (EMORB). Rare earth element (REE) and other immobile trace element characteristics of blueschist assemblage BSI resemble those of normal mid-oceanic ridge basalt (NMORB). These three assemblages are likely formed on a seamount setting, and the prevalent presence of carbonate minerals and omphacite quartzite stripes/gobbets suggests ancient pelagic sediments including marls are probably developed upon the basaltic seamount. Whereas the geochemical characteristics of BS2 assemblage are of volcanic arc basalt-type. The seamount with the pelagic sediments on it is brought into the subduction zone, and volcanic arc basalts formed on the active continental marg...

There are short cycle, intensity disturbance and other characteristics about formation of subsided wetlands in mining coal area with higher groundwater table, nthere are also differences to general wetlands about conversion of i§carbon sources, carbon sinksi¨ and succession law, it has been a focus of attention about coal mining area of ecological restoration. The article aims to reveal the spatial and temporal changes characteristics, control elements of i§carbon pooli¨ in subsided wetlands, to reveal succession rules of i§carbon sinksi¨ in subsided coal mining wetlands with higher groundwater table. To present based target of ecological restoration with carbon of i§adding sinks, source reductioni¨ in mining coal subsided wetlands with higher groundwater table.

Coal mining not only destroyed the ecological balance and the output capacity of the land, but also seriously disturbed the balance of effects of mining of the carbon. Geographical conditions, climate, coal mining methods, ecological damage are the major factors which effect the coal mining areas in the carbon. Coal mining areas from a macro perspective by considering the carbon-carbon effects of the type required influence factors. Coal mining induced land use / cover a larger range of the disturbance, the individual effects of mining division of Micro carbon based on coal mining can be caused by changes in land type, for the coal mining area of the carbon effects of such mechanisms provide the basis for evolution.