Multi-Resolution Learning and Semantic Edge Enhancement for Super-Resolution Semantic Segmentation of Urban Scene Images.

Super-resolution semantic segmentation (SRSS) is a technique that aims to obtain high-resolution semantic segmentation results based on resolution-reduced input images. SRSS can significantly reduce computational cost and enable efficient, high-resolution semantic segmentation on mobile devices with limited resources. Some of the existing methods require modifications of the original semantic segmentation network structure or add additional and complicated processing modules, which limits the flexibility of actual deployment. Furthermore, the lack of detailed information in the low-resolution input image renders existing methods susceptible to misdetection at the semantic edges. To address the above problems, we propose a simple but effective framework called multi-resolution learning and semantic edge enhancement-based super-resolution semantic segmentation (MS-SRSS) which can be applied to any existing encoder-decoder based semantic segmentation network. Specifically, a multi-resolution learning mechanism (MRL) is proposed that enables the feature encoder of the semantic segmentation network to improve its feature extraction ability. Furthermore, we introduce a semantic edge enhancement loss (SEE) to alleviate the false detection at the semantic edges. We conduct extensive experiments on the three challenging benchmarks, Cityscapes, Pascal Context, and Pascal VOC 2012, to verify the effectiveness of our proposed MS-SRSS method. The experimental results show that, compared with the existing methods, our method can obtain the new state-of-the-art semantic segmentation performance.

A Security Information Transmission Method Based on DHR for Seafloor Observation Network.

A seafloor observation network (SON) consists of a large number of heterogeneous devices that monitor the deep sea and communicate with onshore data centers. Due to the long-distance information transmission and the risk of malicious attacks, ensuring the integrity of data in transit is essential. A cryptographically secure frame check sequence (FCS) has shown great advantages in protecting data integrity. However, the commonly used FCS has a collision possibility, which poses a security risk; furthermore, reducing the encryption calculation cost is a challenge. In this paper, we propose a secure, lightweight encryption scheme for transmitted data inspired by mimic defense from dynamic heterogeneous redundancy theory. Specifically, we use dynamic keys to encrypt a data block and generate multiple encrypted heterogeneous blocks for transmission. These continuously changing encrypted data blocks increase the confusion regarding the original encoded data, making it challenging for attackers to interpret and modify the data blocks. Additionally, the redundant information from the multiple blocks can identify and recover tampered data. Our proposed scheme is suitable for resource-constrained environments where lightweight encryption is crucial. Through experimental demonstrations and analysis methods, we determine the effectiveness of our encryption scheme in reducing computational costs and improving security performance to protect data integrity.

Hyperspectral image reconstruction via patch attention driven network.

Coded aperture snapshot spectral imaging (CASSI) captures 3D hyperspectral images (HSIs) with 2D compressive measurements. The recovery of HSIs from these measurements is an ill-posed problem. This paper proposes a novel, to our knowledge, network architecture for this inverse problem, which consists of a multilevel residual network driven by patch-wise attention and a data pre-processing method. Specifically, we propose the patch attention module to adaptively generate heuristic clues by capturing uneven feature distribution and global correlations of different regions. By revisiting the data pre-processing stage, we present a complementary input method that effectively integrates the measurements and coded aperture. Extensive simulation experiments illustrate that the proposed network architecture outperforms state-of-the-art methods.

Low-threshold lasing from bound states in the continuum with dielectric metasurfaces.

Bound states in the continuum (BICs) with extremely large quality factors (Q factors) can enhance the light-matter interaction and thus achieve low-threshold lasing. Here, we theoretically propose and experimentally demonstrate the low-threshold lasing at room temperature based on BICs. A threshold of approximately 306.7 W/cm2 (peak intensity) under a 7.5 ns-pulsed optical excitation is presented in an all-dielectric metasurface system consisting of titanium dioxide (TiO2) nanopillars with a dye film. Also, the multimode lasing can be excited by the higher pumping. Our results may find exciting applications in on-chip coherent light sources, filtering, and sensing.

The branchless gene Clbl in watermelon encoding a TERMINAL FLOWER 1 protein regulates the number of lateral branches.

KEY MESSAGE: A SNP mutation in Clbl gene encoding TERMINAL FLOWER 1 protein is responsible for watermelon branchless. Lateral branching is one of the most important traits, which directly determines plant architecture and crop productivity. Commercial watermelon has the characteristics of multiple lateral branches, and it is time-consuming and labor-costing to manually remove the lateral branches in traditional watermelon cultivation. In our present study, a lateral branchless trait was identified in watermelon material WCZ, and genetic analysis revealed that it was controlled by a single recessive gene, which named as Clbl (Citrullus lanatus branchless). A bulked segregant sequencing (BSA-seq) and linkage analysis was conducted to primarily map Clbl on watermelon chromosome 4. Next-generation sequencing-aided marker discovery and a large mapping population consisting of 1406 F2 plants were used to further map Clbl locus into a 9011-bp candidate region, which harbored only one candidate gene Cla018392 encoding a TERMINAL FLOWER 1 protein. Sequence comparison of Cla018392 between two parental lines revealed that there was a SNP detected from C to A in the coding region in the branchless inbred line WCZ, which resulted in a mutation from alanine (GCA) to glutamate (GAA) at the fourth exon. A dCAPS marker was developed from the SNP locus, which was co-segregated with the branchless phenotype in both BC1 and F2 population, and it was further validated in 152 natural watermelon accessions. qRT-PCR and in situ hybridization showed that the expression level of Cla018392 was significantly reduced in the axillary bud and apical bud in branchless line WCZ. Ectopic expression of ClTFL1 in Arabidopsis showed an increased number of lateral branches. The results of this study will be helpful for better understanding the molecular mechanism of lateral branch development in watermelon and for the development of marker-assisted selection (MAS) for new branchless watermelon cultivars.

Microservice Security Framework for IoT by Mimic Defense Mechanism.

Containers and microservices have become the most popular method for hosting IoT applications in cloud servers. However, one major security issue of this method is that if a container image contains software with security vulnerabilities, the associated microservices also become vulnerable at run-time. Existing works attempted to reduce this risk with vulnerability-scanning tools. They, however, demand an up-to-date database and may not work with unpublished vulnerabilities. In this paper, we propose a novel system to strengthen container security from unknown attack using the mimic defense framework. Specifically, we constructed a resource pool with variant images and observe the inconsistency in execution results, from which we can identify potential vulnerabilities. To avoid continuous attack, we created a graph-based scheduling strategy to maximize the randomness and heterogeneity of the images used to replace the current images. We implemented a prototype using Kubernetes. Experimental results show that our framework makes hackers have to send 54.9% more random requests to complete the attack and increases the defence success rate by around 8.16% over the baseline framework to avoid the continuous unknown attacks.

Transcriptome regulation of carotenoids in five flesh-colored watermelons (Citrullus lanatus).

BACKGROUND: Fruit flesh color in watermelon (Citrullus lanatus) is a great index for evaluating the appearance quality and a key contributor influencing consumers' preferences. But the molecular mechanism of this intricate trait remains largely unknown. Here, the carotenoids and transcriptome dynamics during the fruit development of cultivated watermelon with five different flesh colors were analyzed. RESULTS: A total of 13 carotenoids and 16,781 differentially expressed genes (DEGs), including 1295 transcription factors (TFs), were detected in five watermelon genotypes during the fruit development. The comprehensive accumulation patterns of carotenoids were closely related to flesh color. A number of potential structural genes and transcription factors were found to be associated with the carotenoid biosynthesis pathway using comparative transcriptome analysis. The differentially expressed genes were divided into six subclusters and distributed in different GO terms and metabolic pathways. Furthermore, we performed weighted gene co-expression network analysis and predicted the hub genes in six main modules determining carotenoid contents. Cla018406 (a chaperone protein dnaJ-like protein) may be a candidate gene for ß-carotene accumulation and highly expressed in orange flesh-colored fruit. Cla007686 (a zinc finger CCCH domain-containing protein) was highly expressed in the red flesh-colored watermelon, maybe a key regulator of lycopene accumulation. Cla003760 (membrane protein) and Cla021635 (photosystem I reaction center subunit II) were predicted to be the hub genes and may play an essential role in yellow flesh formation. CONCLUSIONS: The composition and contents of carotenoids in five watermelon genotypes vary greatly. A series of candidate genes were revealed through combined analysis of metabolites and transcriptome. These results provide an important data resource for dissecting candidate genes and molecular basis governing flesh color formation in watermelon fruit.

An integrated transcriptome and metabolome approach reveals the accumulation of taste-related metabolites and gene regulatory networks during watermelon fruit development.

MAIN CONCLUSION: Accumulation patterns and gene regulatory networks of sugars and cucurbitacins and related primary and secondary metabolites during cultivated watermelon 'Cheng Lan' and wild watermelon 'PI 632,751' fruit development were identified. Metabolites are the end products of cellular regulatory processes and play important roles in fruit taste formation. However, comprehensive studies on the accumulation patterns of watermelon fruit metabolites and transcriptional regulatory networks are still scarce. In this study, 451 annotated metabolites were identified at four key fruit developmental stages in wild watermelon 'PI 632,751' and modern cultivated watermelon 'Cheng Lan'. Interestingly, 11 sugars and 25 major primary metabolites were mainly accumulated in 'Cheng Lan' during fruit development, which are considered to be the potential metabolites beneficial to the formation of watermelon taste. Cucurbitacins and the main flavonoids were mainly specifically accumulated in 'PI 632,751', not being considered to be responsible for the taste. Moreover, forty-seven genes involved in carbohydrate metabolism, glycolysis, and TCA cycle were highly expressed in 'Cheng Lan', which was positively correlated with the accumulation of major primary metabolites. Alternatively, seven UDP-glycosyltransferase genes are closely related to the glycosylation of cucurbitacins through co-expression analysis. Our findings established a global map of metabolite accumulation and gene regulation during fruit development in wild and cultivated watermelons and provided valuable information on taste formation in watermelon fruit.

Nonlinear reconstruction of coded spectral X-ray CT based on material decomposition.

Coded spectral X-ray computed tomography (CT) based on K-edge filtered illumination is a cost-effective approach to acquire both 3-dimensional structure of objects and their material composition. This approach allows sets of incomplete rays from sparse views or sparse rays with both spatial and spectral encoding to effectively reduce the inspection duration or radiation dose, which is of significance in biological imaging and medical diagnostics. However, reconstruction of spectral CT images from compressed measurements is a nonlinear and ill-posed problem. This paper proposes a material-decomposition-based approach to directly solve the reconstruction problem, without estimating the energy-binned sinograms. This approach assumes that the linear attenuation coefficient map of objects can be decomposed into a few basis materials that are separable in the spectral and space domains. The nonlinear problem is then converted to the reconstruction of the mass density maps of the basis materials. The dimensionality of the optimization variables is thus effectively reduced to overcome the ill-posedness. An alternating minimization scheme is used to solve the reconstruction with regularizations of weighted nuclear norm and total variation. Compared to the state-of-the-art reconstruction method for coded spectral CT, the proposed method can significantly improve the reconstruction quality. It is also capable of reconstructing the spectral CT images at two additional energy bins from the same set of measurements, thus providing more spectral information of the object.

A rapid and non-invasive fluorescence method for quantifying coenzyme Q10 in blood and urine in clinical analysis.

BACKGROUND: Coenzyme Q10 (CoQ10) supplementation can improve cognition in patients with Alzheimer's disease (AD) and AD transgenic model mice. To ameliorate the discomfort that patients with AD suffer after several blood extractions, a non-invasive method for detecting urine CoQ10 levels needs to be established. METHODS: Here, we developed a new technique of fluorescence spectrophotometry with ethyl cyanoacetate (FS-ECA), on the basis of the principle that the chemical derivative obtained from the interaction between CoQ10 and ECA was detected by a fluorescence detector at λex/em  = 450/515 nm. As a standard reference method, the same batches of the clinical samples were analyzed by high-performance liquid chromatography with an ultraviolet detector (HPLC-UV) at 275 nm. RESULTS: The limits of detection (LOD) and limits of quantization (LOQ) (serum: 0.021 and 0.043 mg/L; urine: 0.012 and 0.025 mg/L) determined by the FS-ECA method were similar to that obtained through HPLC-UV (serum: 0.017 and 0.035 mg/L; urine: 0.012 and 0.025 mg/L). More importantly, this new FS-ECA technique as well as the conventional HPLC-UV method could detect a marked difference in urine CoQ10 levels between AD and controls. CONCLUSION: Our findings suggest that this non-invasive method for quantifying urine CoQ10 potentially replaces HPLC to detect blood CoQ10.

What Determines Good Rehabilitation Recovery in Stroke Patients with very Severe Disability? A Retrospective Cohort Study.

OBJECTIVE: This study aimed to evaluate predictors of good recovery after rehabilitation in stroke patients with very severe disability. METHODS: 168 Eligible patients were enrolled. Fugl-Meyer of motor score (FMM), National Institute of Health Stroke Scale (NIHSS), and Modified Barthel Index (MBI) were used to measure the motor and degree of disability. Logistic regressions were used to investigate the effect of 21 pre-existing conditions on efficacy outcomes after 4 weeks rehabilitation. RESULTS: The results indicated the following: 1) FMM (r = 0.2, P=0.02) and NIHSS (r = -0.2, P=0.00) were significantly correlated (positively and negatively, respectively) with fat-free mass index (FFMI); 2)FFMI (odds ratio [OR],1.9; 95% CI, 1.1 to 3.3; P=0.02), OAI (OR, 0.9; 95% CI, 0.9 to 1.0; P=0.00), and FMM (OR, 1.0; 95% CI, 1.0 to 1.1; P=0.00) were revealed as predictors of good recovery after 4 weeks rehabilitation; 3) The model could predict the possibility of good recovery in stroke patients with very severe disability with excellent sensitivity and specificity(80.0% and 67.1%, respectively). CONCLUSIONS: The results could enable clinicians to predict good recovery after rehabilitation in patients with very severe post-stroke disability.

Nonlinear compressive inverse lithography aided by low-rank regularization.

Photolithography is at the core of the semiconductor industry that is used to fabricate microscale and nanoscale integrated circuits. Inverse lithography is a technique extensively used to compensate for lithography patterning distortions. It refers to methods that pre-distort the photomask patterns such that their projection, through the photolithography system, results in a pattern that is as close as possible to the intended original. However, most inverse lithography technique (ILT) methods suffer from large computational complexity. This paper develops a nonlinear compressive sensing framework for ILT that effectively improves the computational efficiency and image fidelity, while at the same time controlling the mask complexity. Based on a nonlinear lithography imaging model, the compressive ILT is formulated as an inverse optimization problem aimed at reducing the patterning error, and enforcing the sparsity and low rank properties of the mask pattern. A downsampling strategy is adopted to reduce the dimensionality of the cost function, thus alleviating the computational burden. Sparsity and low-rank regularizations are then used to constrain the solution space and reduce the mask complexity. The split Bregman algorithm is used to solve for the inverse optimization problem. The superiority of the proposed method is verified by a set of simulations and comparison to traditional ILT algorithms.

Genetic mapping and development of molecular markers for a candidate gene locus controlling rind color in watermelon.

KEY MESSAGE: ClCG08G017810 (ClCGMenG) encoding a 2-phytyl-1,4-beta-naphthoquinone methyltransferase protein is associated with formation of dark green versus light green rind color in watermelon. Rind color is an important agronomic trait in watermelon [Citrullus lanatus (Thunb.) Matsum. and Nakai], but the underlying molecular mechanism for this trait is not fully known. In the present study, we identified a single locus on chromosome 8 accounting for watermelon rind color (dark green vs. light green). Genetic analysis of F1, F2, and BC1 populations derived from two parental lines (9904 with dark green rind and Handel with light green rind) revealed that the watermelon rind color (dark green vs. light green) is controlled by a single locus, and dark green is dominant to light green rind. Initial mapping revealed a region of interest spanning 2.07 Mb on chromosome 8. Genetic mapping with CAPS and SNP markers narrowed down the candidate region to 31.4 kb. Gene annotation of the corresponding region in the reference genome revealed the ClCG08G017810 gene sequence encoding the 2-phytyl-1,4-beta-naphthoquinone methyltransferase protein. The sequence alignment of the candidate gene with the two parental lines suggested a nonsynonymous SNP mutation in the coding region of ClCG08G017810, converting an arginine (R) to glycine (G). The SNP might be associated with rind color of 103 watermelon germplasm lines investigated in this study. The qRT-PCR analysis revealed higher expression of ClCG08G017810 in dark green rind than in light green rind. Therefore, ClCG08G017810 is a candidate gene associated with watermelon rind color. The present study facilitates marker-assisted selection useful for the development of cultivars with desirable rind color.

Molecular Mapping and Candidate Gene Analysis for GA3 Responsive Short Internode in Watermelon (Citrullus lanatus).

Plants with shorter internodes are suitable for high-density planting, lodging resistance and the preservation of land resources by improving yield per unit area. In this study, we identified a locus controlling the short internode trait in watermelon using Zhengzhouzigua (long internode) and Duan125 (short internode) as mapping parents. Genetic analysis indicated that F1 plants were consistent with long internode plants, which indicates that the long internode was dominant over the short internode. The observed F2 and BC1 individuals fitted the expected phenotypic segregation ratios of 3:1 and 1:1, respectively. The locus was mapped on chromosome 9 using a bulked segregant analysis approach. The region was narrowed down to 8.525 kb having only one putative gene, Cla015407, flanking by CAPS90 and CAPS91 markers, which encodes gibberellin 3ß-hydroxylase (GA 3ß-hydroxylase). The sequence alignment of the candidate gene between both parents revealed a 13 bp deletion in the short internode parent, which resulted in a truncated protein. Before GA3 application, significantly lower GA3 content and shorter cell length were obtained in the short internode plants. However, the highest GA3 content and significant increase in cell length were observed in the short internode plants after exogenous GA3 application. In the short internode plants, the expression level of the Cla015407 was threefold lower than the long internode plants in the stem tissue. In general, our results suggested that Cla015407 might be the candidate gene responsible for the short internode phenotype in watermelon and the phenotype is responsive to exogenous GA3 application.

Genetic mapping reveals a candidate gene (ClFS1) for fruit shape in watermelon (Citrullus lanatus L.).

KEY MESSAGE: A 159 bp deletion in ClFS1 gene encoding IQD protein is responsible for fruit shape in watermelon. Watermelon [Citrullus lanatus (Thunb.) Matsum. & Nakai] is known for its rich diversity in fruit size and shape. Fruit shape has been one of the major objectives of watermelon breeding. However, the candidate genes and the underlying genetic mechanism for such an important trait in watermelon are unknown. In this study, we identified a locus on chromosome 3 of watermelon genome controlling fruit shape. Segregation analysis in F2 and BC1 populations derived from a cross between two inbred lines "Duan125" (elongate fruit) and "Zhengzhouzigua" (spherical fruit) suggests that fruit shape of watermelon is controlled by a single locus and elongate fruit (OO) is incompletely dominant to spherical fruit (oo) with the heterozygote (Oo) being oval fruit. GWAS profiles among 315 accessions identified a major locus designated on watermelon chromosome 3, which was confirmed by BSA-seq mapping in the F2 population. The candidate gene was mapped to a region 46 kb on chromosome 3. There were only four genes present in the corresponding region in the reference genome. Four candidate genes were sequenced in this region, revealing that the CDS of Cla011257 had a 159 bp deletion which resulted in the omission of 53 amino acids in elongate watermelon. An indel marker was derived from the 159 bp deletion to test the F2 population and 105 watermelon accessions. The results showed that Cla011257 cosegregated with watermelon fruit shape. In addition, the Cla011257 expression was the highest at ovary formation stage. The predicted protein of the Cla011257 gene fitted in IQD protein family which was reported to have association with cell arrays and Ca2+-CaM signaling modules. Clear understanding of the genes facilitating the fruit shape along with marker association selection will be an effective way to develop new cultivars.

Construction of A High-Density Genetic Map and Mapping of Fruit Traits in Watermelon (Citrullus Lanatus L.) Based on Whole-Genome Resequencing.

Watermelon (Citrullus lanatus L.) is an important horticultural crop that is grown worldwide and has a high economic value. To dissect the loci associated with important horticultural traits and to analyze the genetic and genomic information of this species, a high-density genetic map was constructed based on whole-genome resequencing (WGR), a powerful high-resolution method for single-nucleotide polymorphism (SNP) marker development, genetic map construction, and gene mapping. Resequencing of both parental lines and 126 recombinant inbred lines (RIL) resulted in the detection of 178,762 single-nucleotide polymorphism (SNP) markers in the parental lines at a sequencing depth greater than four-fold. Additionally, 2132 recombination bin markers comprising 103,029 SNP markers were mapped onto 11 linkage groups (LGs). Substantially more SNP markers were mapped to the genetic map compared with other recent studies. The total length of the linkage map was 1508.94 cM, with an average distance of 0.74 cM between adjacent bin markers. Based on this genetic map, one locus for fruit bitterness, one locus for rind color, and one locus for seed coat color with high LOD scores (58.361, 18.353, 26.852) were identified on chromosome 1, chromosome 8, and chromosome 3, respectively. These prominent loci were identified in a region of 6.16 Mb, 2.07 Mb, and 0.37 Mb, respectively. On the basis of current research, the high-density map and mapping results will provide a valuable tool for identifying candidate genes, map-based gene cloning, comparative mapping, and marker-assisted selection (MAS) in watermelon breeding.

γ-Glutamyl Transferase as a Risk Factor for All-Cause or Cardiovascular Disease Mortality Among 5912 Ischemic Stroke.

BACKGROUND AND PURPOSE: The aim of the study was to evaluate the association of the measurement of serum γ-glutamyl transferase (GGT) concentrations at admission with 1-year all-cause or cardiovascular disease (CVD) mortality in patients with acute ischemic stroke. METHODS: This prospective, multicenter cohort study was conducted in 4 stroke centers in China. Baseline GGT measurements were tested. The relationship of GGT to the risk of death from all-cause or CVD was examined among 1-year follow-up patients. RESULTS: We recorded results from 5912 patients with stroke. In those patients, 51.0% were men, and the median age was 61 years. In both men and women, high GGT was significantly associated with total mortality from all-cause or CVD (P<0.001). The elevated GGT revealed adjusted hazard ratios (95% confidence interval) of 3.03 (1.99-4.54) and 3.24 (2.14-4.92) for mortality from all-cause and CVD, respectively. With an area under the curve of 0.69 (95% confidence interval, 0.66-0.73), GGT showed a significantly greater discriminatory ability to predict all-cause mortality as compared with others factors. GGT improved the National Institutes of Health Stroke Scale score (area under the curve of the combined model, 0.75 [95% confidence interval, 0.73-0.78]; P<0.01). CONCLUSIONS: This study demonstrates that GGT is independently associated with all-cause and CVD mortality in patients with ischemic stroke.

Circulating FABP4 (Fatty Acid-Binding Protein 4) Is a Novel Prognostic Biomarker in Patients With Acute Ischemic Stroke.

BACKGROUND AND PURPOSE: FABP4 (fatty acid-binding protein 4) is an intracellular lipid chaperone involved in coordination of lipid transportation and atherogenesis. This study aimed at observing the effect of FABP4 on the 3-month outcomes in Chinese patients with acute ischemic stroke. METHODS: In a prospective multicenter observational study, serum concentrations of FABP4 were on admission measured in plasma of 737 consecutive patients with acute ischemic stroke. Serum concentrations of FABP4, National Institutes of Health Stroke Scale score, and conventional risk factors were evaluated to determine their value to predict functional outcome and mortality within 3 months. RESULTS: During follow-up, an unfavorable functional outcome was found in 260 patients (35.3%), and 94 patients (12.8%) died. In multivariate models comparing the third and fourth quartiles to the first quartile of FABP4, the concentrations of FABP4 were associated with poor functional outcome and mortality. Compared with the reference category (Q1-Q3), the concentrations of FABP4 in Q4 had a relative risk of 4.77 (95% confidence interval [CI], 2.02-8.15; P<0.001) for poor functional outcome and mortality (odds ratio, 6.15; 95% CI, 3.43-12.68) after adjusting for other significant outcome predictors in univariate logistic regression analysis. Receiver-operating characteristic curves to predict poor functional outcome and mortality demonstrated areas under the curve of FABP4 of 0.78 (95% CI, 0.75-0.82) and 0.83 (95% CI, 0.79-0.88), which improved the prognostic accuracy of National Institutes of Health Stroke Scale score with combined areas under the curve of 0.83 (95% CI, 0.76-0.89; P<0.01) and 0.86 (95% CI, 0.81-0.92), respectively. CONCLUSIONS: Data show that FABP4 is a novel independent prognostic marker improving the currently used risk stratification of stroke patients.

Fast lithography aerial image calculation method based on machine learning.

Aerial image calculation for thick masks is an indispensable but time-consuming step in most lithography simulations. This paper develops a fast thick-mask aerial image calculation method based on machine learning for partially coherent lithography systems. First, some sparse sampling points are chosen from the source plane to represent the partially coherent illumination. Then, the training libraries of thick-mask diffraction near-fields are built up for all sampling points based on a set of representative mask features. For an arbitrary thick mask, we calculate its aerial image using the nonparametric kernel regression technique and the pre-calculated training libraries. Subsequently, a post-processing method is applied to compensate for the estimation error and improve the computational accuracy. In addition, this paper also studies the impacts of several key factors on the accuracy and efficiency of the proposed method. Finally, the proposed method is verified by the simulations at 45 nm and 14 nm technology nodes.

Serum 25-hydroxyvitamin D predicts the short-term outcomes of Chinese patients with acute ischaemic stroke.

Low vitamin D levels have been reported to contribute to the risk of cardiovascular events and mortality, especially stroke. In the present study we therefore evaluated the short-term prognostic value of serum 25(OH)D (25-hydroxyvitamin D) in Chinese patients with AIS (acute ischaemic stroke). From February 2010 to September 2012, consecutive stroke patients admitted to the emergency department at two hospitals in Beijing, China were identified. Clinical information was collected, and the serum concentration of 25(OH)D and NIHSS (National Institutes of Health Stroke Scale) were measured at the time of admission. Short-term functional outcome was measured using a modified Rankin Scale (mRS) at 90 days after admission. Multivariate analyses were performed using logistic regression models. During the inclusion period, 231 patients were diagnosed as having AIS, and 220 completed follow-up. The median serum 25(OH)D level was significantly lower in patients with AIS compared with normal controls [14.2 (10.2-18.9) ng/ml compared with 17.9 (12.5-22.9) ng/ml; P<0.001; values are medians (interquartile range)]. 25(OH)D was an independent prognostic marker of short-term functional outcome and death {0.79 (0.73-0.85) and 0.70 (0.50-0.98) respectively [values are odds rations (95% confidence intervals)]; P<0.01 for both, adjusted for NHISS, other predictors and vascular risk factors} in patients with AIS. In ROC (receiver operating characteristic) curve analysis, the prognostic accuracy of 25(OH)D was higher compared with all of the other serum predictors and was in the range of NIHSS score. In conclusion, these findings suggest that 25(OH)D is an independent prognostic marker for death and functional outcome within 90 days in Chinese patients with AIS even after adjusting for possible confounding factors.