Modified UNet Model for Brain Stroke Lesion Segmentation on Computed Tomography Images

The task of segmentation of brain regions affected by ischemic stroke is help to tackle important challenges of modern stroke imaging analysis. Unfortunately, at the moment, the models for solving this problem using machine learning methods are far from ideal. In this paper, we consider a modified 3D UNet architecture to improve the quality of stroke segmentation based on 3D computed tomography images. We use the ISLES 2018 (Ischemic Stroke Lesion Segmentation Challenge 2018) open dataset to train and test the proposed model. Interpretation of the obtained results, as well as the ideas for further experiments are included in the paper. Our evaluation is performed using the Dice or f1 score coefficient and the Jaccard index. Our architecture may simply be extended to ischemia segmentation and computed tomography image identification by selecting relevant hyperparameters. The Dice/f1 score similarity coefficient of our model shown 58% and results close to ground truth which is higher than the standard 3D UNet model, demonstrating that our model can accurately segment ischemic stroke. The modified 3D UNet model proposed by us uses an efficient averaging method inside a neural network. Since this set of ISLES is limited in number, using the data augmentation method and neural network regularization methods to prevent overfitting gave the best result. In addition, one of the advantages is the use of the Intersection over Union loss function, which is based on the assessment of the coincidence of the shapes of the recognized zones.     

基于多策略的乌孜别克语名词词干识别研究

乌孜别克语名词词干识别是自然语言处理领域的基础研究,主要方法是从句子中提取名词词干,提高名词标注效率和准确性。该文首先陈述形态分析、形态特征对识别其词性的作用,然后讨论乌孜别克语的词类划分标准、名词的形态特征,乌孜别克语西尔里文转换拉丁文,乌孜别克语词汇翻译、标注技术,总结词缀歧义及消解规则。该文提出利用形态规则、词典、最大熵融合策略,设计现代乌孜别克语新词中名词词干识别算法,其中包括特征选择及参数估计、词内部特征、前后依存词特征等。最后以乌孜别克语网站文本作为验证对象,对名词词干进行统计与分析。     

Temperature variation in high slump drilled shaft concrete and its effect on slump loss

Drilled shaft refers to a deep foundation system where a single large diameter pier is used to replace a whole group of piles. High slump self-compacting concrete is used in drilled shafts due to its high fluidity and less proneness to segregation. The Florida Department of Transportation (FDOT) specifications require that such concrete should have a slump between 7 and 9 in. when placed and should maintain a slump of 4 in. or more throughout the concrete placement time. Furthermore, the mix for the slump loss test should be prepared at a temperature consistent with the highest ambient or initial concrete temperature (whichever is greater) expected during actual concrete placement. It is possible that the temperature of concrete inside the drilled shaft is lower than the ambient or initial concrete temperature due to the presence of ground water. If that is the case, slump loss would be less than the loss determined at the highest ambient or the initial concrete temperature, making the FDOT requirement unrealistic. However, it should be experimentally verified. This experimental study was conducted with the objective to establish profiles of concrete temperature in time from placement to hardening along depth as well as across width of the drilled shaft. Based on the gathered data, it was found that no significant temperature differential existed along the depth and across the width of the drilled shaft during the initial setting of concrete. The temperature of concrete inside the drilled shaft was same as initial concrete temperature before placement at all locations. This finding leads to the conclusion that concrete temperature inside drilled shaft is not affected by ambient temperature and/or the underground temperature conditions.     

Counterion-Induced Cationic Flexible Nanoparticle Catalytic of Piperidinolysis of Phenyl Salicylate

The removal of PSa^? from bulk aqueous phase to the pseudo‐micellar phase by halobenzoate counterion X is responsible for the monotonic increase in k_obs (pseudo first‐order rate constants) with the increase in the values of [MX] where MX = sodium salts of 2‐, and 4‐halobenzoic acids. The values of ion exchange constants, or for X = 2‐ and 4‐halobenzoate ions in the presence of tetradecyltrimethylammonium bromide (TTABr) were calculated from the apparent catalytic rate constants, ^Xk_cat which represent the catalytic effect of CFN. Larger values of or were observed for X = 4‐halobenzoate ions than that for X = 2‐halobenzoate ions due to isomeric factors. The values of or determined in the presence of TTABr were compared with previously determined or values in the presence of cetyltrimethylammonium bromide (CTABr). The values of or are nearly 8 ~ 9‐fold larger for 4‐IBz^?, 4‐BrBz^? and 4‐ClBz^? compared to the respective values of X = 2‐IBz^?, 2‐BrBz^? and 2‐ClBz^?. The values of or for X = 4‐FBz^? is nearly 3‐fold larger than that for X = 2‐FBz^?. The values of or for X = 2‐ and 4‐halobenzoates are significantly smaller in the presence of TTABr than these in the presence of CTABr nanoparticles.     

Deterioration and Recovery of Metakaolin Blended Concrete Subjected to High Temperature

The deterioration and spalling frequency of metakaolin (MK) blended concrete subjected to high temperature is analyzed and compared with the equivalent silica fume (SF), fly ash (FA) and pure OPC concretes. Normal and high strength concrete mixes incorporating 0–20% MK were prepared and exposed to a series of high temperatures till 800°C. The residual compressive strength, porosity and pore size distribution were determined. It was found that after an increase in compressive strength at 200°C, the MK concrete suffered a more severe loss of compressive strength and permeability-related durability than the corresponding SF, FA, and pure OPC concretes at higher temperatures. Explosive spalling was observed in both normal and high strength MK concretes, and the frequency increased with higher MK contents. It is concluded that the dense micro-structure and low porosity are the main reasons of the poor performance of MK concrete at high temperature. The effect of post-fire curing on the strength and durability recovery of fire-damaged concrete was also investigated. The test results indicated that the post-fire curing, results in substantial strength and durability recovery, and its extent depends upon the types of concrete, exposure temperature, method, and duration of re-curing.     

Measurements of 222Rn activity concentration in domestic water sources in Penang, northern peninsular Malaysia

Measurements of (222)Rn activity concentration were carried out in 39 samples collected from the domestic and drinking water sources used in the island and mainland of Penang, northern peninsular, Malaysia. The measured activity concentrations ranged from 7.49 to 26.25 Bq l(-1), 0.49 to 9.72 Bq l(-1) and 0.58 to 2.54 Bq l(-1) in the raw, treated and bottled water samples collected, respectively. This indicated relatively high radon concentrations compared with that from other parts of the world, which still falls below the WHO recommended treatment level of 100 Bq l(-1). From this data, the age-dependent associated committed effective doses due to the ingestion of (222)Rn as a consequence of direct consumption of drinking water were calculated. The committed effective doses from (222)Rn resulting from 1 y's consumption of these water were estimated to range from 0.003 to 0.048, 0.001 to 0.018 and 0.002 to 0.023 mSv y(-1), for age groups 0-1, 2-16 and >16 y, respectively.     

Large-energy-shift photon upconversion in degenerately doped InP nanowires by direct excitation into the electron gas

Realizing photon upconversion in nanostructures is important for many next- generation applications such as biological labelling, infrared detectors and solar cells. In particular nanowires are attractive for optoelectronics because they can easily be electrically contacted. Here we demonstrate photon upconversion with a large energy shift in highly n-doped InP nanowires. Crucially, the mechanism responsible for the upconversion in our system does not rely on multi-photon absorption via intermediate states, thus eliminating the need for high photon fluxes to achieve upconversion. The demonstrated upconversion paves the way for utilizing nanowires--with their inherent flexibility such as electrical contactability and the ability to position individual nanowires--for photon upconversion devices also at low photon fluxes, possibly down to the single photon level in optimised structures.