Search

The full-bridge can further improve the output power through parallel inverters. The inverter output voltage of single-switch LC resonant circuit is a combination of trapezoidal wave and half sine wave, there is no relevant research on parallel connection of single-switch LC resonant circuit. Therefore, this paper presents a novel input-parallel single-switch LC resonant circuit, this topology adopts switchable secondary networks for constant-current (CC) and constant-voltage (CV) output, and prevents shoot-through of power switches. The decoupling between transmitters simplifies the analysis and calculation. The input-parallel structure avoids the unbalanced input voltage of each inverter and improves the stability of the WPT system. CC and CV output modes can be achieved by controlling one relay without adding DC-DC converters or changing switching frequency. This paper includes decoupling equivalent analysis of coils, topologies analysis of CC mode and CV mode, calculation of equivalent input AC voltage source, design of magnetic coupler and circuit parameters. Finally, a 1-kW experimental prototype is built to verify the theoretical analysis, the maximum efficiency can reach 92.5%.

Analyzing short texts infers discriminative and coherent latent topics that is a critical and fundamental task since many real-world applications require semantic understanding of short texts. Traditional long text topic modeling algorithms (e.g., PLSA and LDA) based on word co-occurrences cannot solve this problem very well since only very limited word co-occurrence information is available in short texts. Therefore, short text topic modeling has already attracted much attention from the machine learning research community in recent years, which aims at overcoming the problem of sparseness in short texts. In this survey, we conduct a comprehensive review of various short text topic modeling techniques proposed in the literature. We present three categories of methods based on Dirichlet multinomial mixture, global word co-occurrences, and self-aggregation, with example of representative approaches in each category and analysis of their performance on various tasks. We develop the first comprehensive open-source library, called STTM, for use in Java that integrates all surveyed algorithms within a unified interface, benchmark datasets, to facilitate the expansion of new methods in this research field. Finally, we evaluate these state-of-the-art methods on many real-world datasets and compare their performance against one another and versus long text topic modeling algorithm., arXiv admin note: text overlap with arXiv:1808.02215 by other authors

Background: Oligodendrocyte precursor cells (OPCs) are myelinated glial cells of the central nervous system (CNS), able to regenerate oligodendrocytes and myelin. This study aimed to elucidate the effect of A2B5-positive (A2B5+) OPC transplantation in rats with spinal cord contusion (SCC) and to investigate changes in expression of various factors involved in the Notch signaling pathway after OPC transplantation. Methods: OPCs were obtained from induced pluripotent stem cells (iPSCs) originating from mouse embryo fibroblasts (MEFs). After identification of iPSCs and iPSC-derived OPCs, A2B5+OPCs were transplanted into the injured site of rats with SCC one week after SCC insult. Behavioral tests evaluated motor and sensory function 7 days after OPC transplantation. Real-time quantitative polymerase chain reaction (RT-qPCR) determined the expression of various cytokines related to the Notch signaling pathway after OPC transplantation. Results: IPSC-derived OPCs were successfully generated from MEFs, as indicated by positive immunostaining of A2B5, PDGFα and NG2. Further differentiation of OPCs was identified by immunostaining of Olig2, Sox10, Nkx2.2, O4, MBP and GFAP. Importantly, myelin formation was significantly enhanced in the SCC+OPC group and SCI-induced motor and sensory dysfunction was largely alleviated by A2B5+OPC transplantation. Expression of factors involved in the Notch signaling pathway (Notch-1, Numb, SHARP1 and NEDD4) was significantly increased after OPC transplantation. Conclusions: A2B5+OPC transplantation attenuates motor and sensory dysfunction in SCC rats by promoting myelin formation, which may be associated with change in expression of factors involved in the Notch signaling pathway.

Hypertension or angiotensin II (Ang II) induces cardiac inflammation and fibrosis, thus contributing to cardiac remodeling. MicroRNAs (miRNAs) are considered crucial regulators of cardiac homeostasis and remodeling in response to various types of stress. It has been reported that miR-451a is involved in regulating ischemic heart injury. However, its role in Ang II-induced cardiac fibrosis remains unknown. Cardiac remodeling was induced in mice by infusion of low-dose Ang II (490 ng/kg/min) with a minipump for 2 weeks. Echocardiography and histological examinations were performed to evaluate cardiac function and pathological changes. We observed that miR-451a expression was the most significantly downregulated in the hearts of Ang II-infused mice and in both primary cardiac myocytes and fibroblasts. Overexpression of miR-451a in mice significantly attenuated Ang II-induced cardiac fibrosis and inflammation. Conversely, knockdown of miR-451a in mice aggravated this effect. Bioinformatics analysis and a luciferase reporter assay revealed that TBX1 was a direct target of miR-451a. Mechanistically, miR-451a directly targeted TBX1 expression, which inhibited TGF-β1 production in both cardiac myocytes and fibroblasts, inactivating of TGF-β1/SMAD2/3 signaling, inhibiting myofibroblast differentiation and proinflammatory cytokine expression, and leading to attenuation of cardiac fibrosis and inflammation. In conclusion, these results indicate that miR-451a acts as a novel regulator of Ang II-induced cardiac fibrosis and inflammation by directly targeting TBX1, and may be a promising therapeutic target for treating hypertensive cardiac diseases.

How to represent and classify a testing sample for the representation-based classification (RBC) plays an important role in the filed of pattern recognition. As a typical kind of the representation-based classification with promising performance, collaborative representation-based classification (CRC) adopts all the training samples to collaboratively represent and then classify each testing sample with the reconstructive residuals among all the classes. However, most of the CRC methods fail to make full use of the localities and discrimination information of data in collaborative representation. To address this issue to further improve the classification performance, we design a novel supervised CRC method entitled locality-constrained weighted collaborative-competitive representation-based classification (LWCCRC). In the proposed method, the localities of data are taken into account by using the positive and negative nearest samples of each testing sample with their corresponding weighted constraints. Such devised locality-constrained weighted term can model the similarity and natural discrimination information contained in the neighborhood region for each testing sample to obtain the favorable representation. Moreover, a competitive constraint is introduced to enhance pattern discrimination among the categorical collaborative representations. To explore the effectiveness of our proposed LWCCRC, the extensive experiments are carried out on three different types of data sets. The experimental results demonstrate that the proposed LWCCRC significantly outperforms the recent state-of-the-art CRC methods.

Zhen-Yi Long,1,* Yi-Feng Zhou,2,* Hao Yuan,1 Ya-Meng Peng,1 Si-Xian Wu,1 Fang Peng1 1Department of Clinical Laboratory, Hunan Provincial People’s Hospital, The First Affiliated Hospital of Hunan Normal University, Changsha, 410005, People’s Republic of China; 2Operating Room, Hunan Provincial People’s Hospital, The First Affiliated Hospital of Hunan Normal University, Changsha, 410005, People’s Republic of China*These authors contributed equally to this workCorrespondence: Hao YuanDepartment of Clinical Laboratory, Hunan Provincial People’s Hospital, The First Affiliated Hospital of Hunan Normal University, Changsha, 410005, People’s Republic of ChinaTel +86-13677366519Fax +86-731-82278012Email yuanhao666@hunnu.edu.cnPurpose: We explored the expression levels of IgG4 and interleukin (IL)-21 in the serum and ankle joints of collagen-induced arthritis (CIA) rats at different disease stages.Materials and Methods: Wistar rats were randomly divided into normal and model groups, and the latter group was administered bovine type II collagen to induce arthritis. Enzyme-linked immunosorbent assay was performed at 21, 28, 35, and 42 days to detect IgG4 and IL-21 in the serum, followed by histological and immunohistochemical analyses of IgG4 and IL-21r in the ankle joint of rats.Results: The contents of IgG4 and IL-21 in the serum of the CIA model group were positively correlated and increased with disease progression. The expression of IgG4 and IL-21 receptors in the ankle joint of the CIA model group was significantly higher than that in the control group. These proteins were closely related to the pathological score. The serum IL-21 level in the model group was closely related to the level of IL-21 receptor in the ankle joint.Conclusion: IL-21 may promote the occurrence and development of rheumatoid arthritis by combining with IL-21r to regulate the content of IgG4.Keywords: rheumatoid arthritis, immunoglobulin, interleukin, interleukin receptor

The emergence of edge computing can effectively tackle the problem of large transmission delays caused by the long-distance between user devices and remote cloud servers. Users can offload tasks to the nearby edge servers to perform computations, so as to minimize the average task response time through effective task dispatching and scheduling methods. However: 1) in the task dispatching phase, the dynamic features of network conditions and server loads make it difficult for the offloaded tasks to select the optimal edge server and 2) in the task scheduling phase, each edge server may face a large number of offloading tasks to schedule, resulting in long average task response time, or even severe task starvation. In this article, we propose an online task dispatching and fair scheduling method OTDS to tackle the above two challenges, which combines online learning (OL) and deep reinforcement learning (DRL) techniques. Specifically, using an OL approach, OTDS performs real-time estimating of network conditions and server loads, and then dynamically assigns tasks to the optimal edge servers accordingly. Meanwhile, at each edge server, by combing the round-robin mechanism with DRL, OTDS is able to allocate appropriate resources to each task according to its time sensitivity and achieve high efficiency and fairness in task scheduling. Evaluation results show that our online method can dynamically allocate network resources and computing resources to those offloaded tasks according to their time-sensitive requirements. Thus, OTDS outperforms the existing methods in terms of the efficiency and fairness on task dispatching and scheduling by significantly reducing the average task response time.

High temperature or temperature nonuniformity has become a serious threat to performance and reliability of high-performance integrated circuits (ICs). Since the temperature in a 3-D IC is mainly determined by a power distribution across tiers, this article proposes a novel method to allocate modules to tiers while exploring a better power distribution according to the hyperparameter optimization technique. In addition, we integrate the sub-3-D thermal mask into the analytical formulation to interleave high power consumption modules in contiguous tiers during distributing modules over placement regions so that it is possible to insert through-silicon vias (TSVs) around high power modules to further reduce the temperature at a later stage. Since the temperature in a tier will be changed every time the locations of TSVs in its lower tier are moved, we also propose a procedure to update the temperature map before refining locations of TSVs. Experimental results have demonstrated that the proposed methodology can effectively reduce the temperature of a 3-D IC with a slight increase in the wirelength. Moreover, its runtime is quite fast.

Objectives Progression of dengue is often associated with thrombocytopenia resulting from viral-induced bone marrow suppression and immune-mediated peripheral platelet consumption. Immature platelet fraction (IPF), which can be measured using a haematology analyser, is a precursor indicating platelet formation in the bone marrow. This study evaluated the trend of IPF as an early recovery indicator of platelets in dengue patients with thrombocytopenia, and its relationship with severe dengue in conjunction with reticulocyte count. Methods Hospitalized patients with dengue were enrolled and followed-up daily until discharge. Blood investigations included daily full blood counts and IPF measured using a haematology analyser. Results In total, 287 patients with confirmed dengue were enrolled in this study, 25 of whom had severe dengue. All patients had a decreasing trend in platelet count in the first week of illness, concomitant with an increasing trend in the percentage of immature platelets to total platelets (IPF%) for more than 3 days prior to platelet recovery. IPF% was significantly increased in patients with severe dengue compared with patients with non-severe dengue on days 3–5 after the onset of fever. Reticulocyte count increased significantly in patients with severe dengue on day 5. Conclusions IPF can be utilized as an early recovery indicator of platelets in patients with dengue and thrombocytopenia.

The outbreak of coronavirus disease 2019 (COVID-19) is a global health emergency. Various omics results have been reported for COVID-19, but the molecular hallmarks of COVID-19, especially in those patients without comorbidities, have not been fully investigated. Here we collect blood samples from 231 COVID-19 patients, prefiltered to exclude those with selected comorbidities, yet with symptoms ranging from asymptomatic to critically ill. Using integrative analysis of genomic, transcriptomic, proteomic, metabolomic and lipidomic profiles, we report a trans-omics landscape for COVID-19. Our analyses find neutrophils heterogeneity between asymptomatic and critically ill patients. Meanwhile, neutrophils over-activation, arginine depletion and tryptophan metabolites accumulation correlate with T cell dysfunction in critical patients. Our multi-omics data and characterization of peripheral blood from COVID-19 patients may thus help provide clues regarding pathophysiology of and potential therapeutic strategies for COVID-19., COVID-19 is a critical public health threat, but molecular characterizations of patients’ immunity is still lacking. Here the authors collected blood from patients with various disease severity, and prefiltered to exclude selected comorbidity, to obtain genomic, transcriptomic, proteomic, metabolomic and lipidomic profiles to report a trans-omics landscape.

Droplet digital loop-mediated isothermal amplification (ddLAMP) is an important assay for pathogen detection due to its high accuracy, specificity, and ability to quantify nucleic acids. However, performing ddLAMP requires expensive instrumentation and the need for highly trained personnel with expertise in microfluidics. To make ddLAMP more accessible, a ddLAMP assay is developed, featuring significantly decreased operational difficulty and instrumentation requirements. The proposed assay consists of three simplified steps: (1) droplet generation step, in which a LAMP mixture can be emulsified just by manually pulling a syringe connected to a microfluidic device. In this step, for the first time, we verify that highly monodispersed droplets can be generated with unstable flow rates or pressures, allowing untrained personnel to operate the microfluidic device and perform ddLAMP assay; (2) heating step, in which the droplets are isothermally heated in a water bath, which can be found in most laboratories; and (3) result analysis step, in which the ddLAMP result can be determined using only a fluorescence microscopy and an open-source analyzing software. Throughout the process, no droplet microfluidic expertise or equipment is required. More importantly, the proposed system enables multiple samples to be processed simultaneously with a detection limit of 10 copies/μL. The test is simple and intuitive to operate in most laboratories for multi-sample detection, significantly enhancing the accessibility and detection throughput of the ddLAMP technique.

Face super-resolution (FSR) is an effective way to deal with low-resolution (LR) face images, which can infer the latent high-resolution (HR) face images from the LR inputs. In contrast with traditional FSR methods such as interpolation, learning-based methods generate more realistic HR images of LR faces by exploiting the relationship between HR and LR images . In this paper, we propose a novel FSR learning approach based on orthonormalized partial least squares referred to as OPLS-SR. It first learns a latent coherent feature space of low-dimensional HR and LR face embeddings via a recursive optimization, and then super-resolves the LR face images through global face reconstruction and facial detail compensation. Experimental results on the CAS-PEAL-R1 and FERET face databases have demonstrated the effectiveness of the proposed OPLS-SR method in terms of quantitative and qualitative evaluations.

Anaerobic-anoxic-oxic (A2O) reactors, as the core parts of wastewater treatment process (WWTP), have attracted considerable attention to achieve the reliability of denitrification and dephosphorization. However, it is difficult to realize the optimal operation of A2O reactors due to the existence of nonlinear dynamics and large uncertainties. To solve this problem, a robust optimal control (ROC) strategy is developed to improve the operation performance of A2O reactors. First, data-driven systematic evaluation criteria are developed to describe the operational indicators of changeable conditions. Second, a robust optimization algorithm is designed to select the optimal solution. Third, a fuzzy neural network (FNN) is used to track the optimal solution in the control process. Finally, this proposed ROC strategy is applied to the phosphorus removal benchmark simulation model (BSM1-P) and the real A2O reactors. The results demonstrate that the strategy developed in this paper has great potential for application in real A2O reactors.

Herpes simplex keratitis (HSK) is a significant cause of vision impairment worldwide. Currently, there are no set diagnostic criteria, and popular diagnostic methods, including clinical examination of the eye via slit lamp examination, could lead to false-negatives and misdiagnoses. Molecular testing with polymerase chain reaction (PCR) may lack concordance with clinical findings, posing a great challenge to ophthalmologists. We evaluate recent studies on techniques for the diagnosis of HSK. We included a total of 23 studies published between 2010 and 2020 in English on diagnostic techniques, including in vivo confocal microscopy, polymerase PCR testing, protein detection in tear film with enzyme-linked immunosorbent assay, and various other protein assays. Although PCR has been widely used as one of the current diagnostic methods for HSK, most studies evaluated its efficacy after including alterations to its normal protocol. Tear sample analysis was performed using multiple tools, although corneal scrapings demonstrated a higher positive detection rate. Diagnostic tools identified were able to detect HSK with varying accuracy. Newer diagnostic techniques like multiplex dot hybridization assay and immunochromatographic assays may be considered as the point-of-care preliminary diagnostic tools. More reliable results may be generated by developing a standardized diagnostic protocol.

BACKGROUND Hepatitis B virus (HBV) remains one of the most serious and prevalent health problems in the world. OBJECTIVES To determine the serum hepatitis B virus (HBV) RNA levels in patients with chronic hepatitis B (CHB) with low HBV DNA levels and analyze the influencing factors. MATERIAL AND METHODS Seventy-two CHB patients with low HBV DNA levels were enrolled and divided into 2 groups according to hepatitis B e antigen (HBeAg) status; their age, sex, the incidence of HBV RNA level < lower limit of detection (LLD), and serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), quantitative determination of HBsAg (qHBsAg), HBV DNA, and HBV RNA levels were compared. The factors influencing serum HBV RNA levels < LLD and the correlation between serum HBV RNA levels, and serum ALT, AST, qHBsAg and HBV DNA levels were analyzed. RESULTS In HBeAg-positive patients, serum AST, qHBsAg and HBV RNA levels were higher, and serum HBV DNA levels and incidence of HBV RNA < LLD were lower than those in HBeAg-negative patients (p < 0.05). Multivariate linear regression analysis revealed that HBeAg is a factor that significantly influences serum HBV RNA levels in patients with CHB (p < 0.05). Multivariate logistic regression analysis indicated that HBeAg and qHBsAg are factors that influence serum HBV RNA levels < LLD in patients with CHB. In HBeAg-positive patients, serum HBV RNA levels were positively correlated with qHBsAg and HBeAg. CONCLUSIONS The serum HBV RNA levels in CHB patients with low HBV DNA levels varied according to HBeAg status. The HBeAg is a factor that significantly influences serum HBV RNA levels in patients with CHB, while HBeAg and qHBsAg are factors that significantly influence serum HBV RNA levels < LLD in patients with CHB.

Amyloid-β (Aβ) peptides and hyperphosphorylated tau protein are the most important pathological markers of Alzheimer’s disease (AD). Neuroinflammation and oxidative stress are also involved in the development and pathological mechanism of AD. Hypoxia inducible factor-1α (HIF-1α) is a transcriptional factor responsible for cellular and tissue adaption to low oxygen tension. Emerging evidence has revealed HIF-1α as a potential medicinal target for neurodegenerative diseases. On the one hand, HIF-1α increases AβPP processing and Aβ generation by promoting β/γ-secretases and suppressing α-secretases, inactivates microglia and reduces their activity, contributes to microglia death and neuroinflammation, which promotes AD pathogenesis. On the other hand, HIF-1α could resist the toxic effect of Aβ, inhibits tau hyperphosphorylation and promotes microglial activation. In summary, this review focuses on the potential complex roles and the future perspectives of HIF-1α in AD, in order to provide references for seeking new drug targets and treatment methods for AD.

We collected neonatal neurological, clinical, and imaging data to study the neurological manifestations and imaging characteristics of neonates with coronavirus disease 2019 (COVID-19). This case–control study included newborns diagnosed with COVID-19 in Wuhan, China from January 2020 to July 2020. All included newborns had complete neurological evaluations and head magnetic resonance imaging. We normalized the extracted T2-weighted imaging data to a standard neonate template space, and segmented them into gray matter, white matter, and cerebrospinal fluid. The comparison of gray matter volume was conducted between the two groups. A total of five neonates with COVID-19 were included in this study. The median reflex scores were 2 points lower in the infected group than in the control group (P = 0.0094), and the median orientation and behavior scores were 2.5 points lower in the infected group than in the control group (P = 0.0008). There were also significant differences between the two groups in the total scale score (P = 0.0426). The caudate nucleus, parahippocampal gyrus, and thalamus had the strongest correlations with the Hammersmith neonatal neurologic examination (HNNE) score, and the absolute correlation coefficients between the gray matter volumes and each part of the HNNE score were all almost greater than 0.5. We first compared the neurological performance of neonates with and without COVID-19 by quantitative neuroimaging and neurological examination methods. Considering the limited numbers of patients, more studies focusing on the structural or functional aspects of the virus in the central nervous system in different age groups will be carried out in the future.

Anterior controllable antedisplacement and fusion (ACAF) is a novel surgical technique for the treatment of ossification of the posterior longitudinal ligament (OPLL). Its prognostic factors for decompression have not been well studied. Additionally, no detailed radiological standard has been set for hoisting the vertebrae-OPLL complex (VOC) in ACAF.To identify the possible prognostic factors for decompression outcomes after ACAF for cervical OPLL, to determine the critical value of radiological parameters for predicting good outcomes, and to establish a radiological standard for hoisting the VOC in ACAF.This was a retrospective multicenter study.A total of 121 consecutive patients with OPLL who underwent ACAF at a point between January 2017 and June 2018 at any one of seven facilities and were monitored for at least 1 year afterward were enrolled in a multicenter study.Japanese Orthopedic Association (JOA) scores, recovery rate (RR) of neurologic function, and surgical complications were used to determine the effectiveness of ACAF.Patients were divided into two groups according to their RR for neurologic function. Patients with an RR of ≥50% and an RR of50% were designated as having good and poor decompression outcomes, respectively. The relationship between various possible prognostic factors and decompression outcomes was assessed by univariate and multivariate analysis. The receiver operating characteristic curve was used to determine the optimal cutoff value of the radiological parameters for prediction of good decompression outcomes. Next, the patients were redivided into three groups according to the cutoff value of the selected radiological parameter (postoperative anteroposterior canal diameter [APD] ratio). Patients with postoperative APD ratios of ≤80.7%, 80.7%-100%, and ≥100% were defined as members of the incomplete, optimal, and excessive antedisplacement groups, respectively. Differences in decompression outcomes among the three groups were compared to verify the reliability of the postoperative APD ratio and assess the necessity of excessive antedisplacement.Multivariate logistic regression analysis showed that patients' age at surgery (odds ratio [OR]=1.18; 95% confidence interval [CI]=1.08-1.29; p.01) and postoperative APD ratio (OR=0.83; 95% CI=0.77-0.90; p.01) were independently associated with decompression outcomes. The optimal cutoff point of the postoperative APD ratio was calculated at 80.7%, with 86.2% sensitivity and 73.5% specificity. There were no significant differences in the postoperative JOA scores and RRs between the excessive antedisplacement group and optimal antedisplacement group (p.05). However, a lower incidence of cerebrospinal fluid leakage and screw slippage was observed in the optimal antedisplacement group (p.05).Patients' age at surgery and their postoperative APD ratio are the two prognostic factors of decompression outcomes after ACAF. The postoperative APD ratio is also the most accurate radiological parameter for predicting good outcomes. Our findings suggest that it is essential for neurologic recovery to restore the spinal canal to more than 80.7% of its original size (postoperative APD ratio80.7%), and restoration to less than its original size (postoperative APD ratio100%) will help reduce the incidence of surgical complications. This may serve as a valuable reference for establishment of a radiological standard for hoisting the VOC in ACAF.

Photocatalytic water splitting has always been a field where breakthroughs are expected to solve energy and environmental problems. However, current catalysts suffer from low activity in mismatched catalytic environments and high cost. Herein, we designed a series of integrated CsPbBr3-CsPbCl3 heterostructures to explore their catalytic capability. Based on extensive calculations, we discovered the inner connection between dopant atoms and the catalytic performance and proposed a new descriptor by applying the Least Absolute Shrinkage and Selection Operator (LASSO) analysis. After systematic screening, the CsPbBr3:Ni-CsPbCl3:Co system is found to be promising for single-catalyst overall water splitting under the same environment. Furthermore, a smaller bandgap that covers the redox potential of water splitting suggests the capability for photocatalysis. Besides, the CsPbBr3:Ni-CsPbCl3:Co system bulk-doped by Co could conduct the photocatalysis with better performance.

Multiple myeloma (MM), a clonal plasma cell malignancy, composes around 10% of hematologic malignancies. Though recent advances in treatment have dramatically improved MM survival, some aggressive courses of disease and dismal outcomes still exist. Low body weight, undernutrition, and cachexia are noted at MM diagnosis. We aim to evaluate the impact of low body mass index (BMI) and undernutrition in MM patients. We retrospectively analyzed MM patients at Taipei Veterans General Hospital in Taiwan between January 1, 2006 and October 31, 2018. Being underweight is defined as having a BMI of under 18.5 kg/m2. The patient’s baseline characteristics, including BMI, serum albumin level, and comorbidities, etc., were recorded. The primary endpoint of the study was all-cause mortality. A Cox regression model was used to estimate the risk factors of mortality. A total of 378 newly diagnosed MM patients were enrolled in this study. The median age of the patients was 69. Thirty patients (7.9%) were underweight at diagnosis. The median overall survival was 1.3 years (95% CI 0.3–5.7) and 5.0 years (95% CI 3.1–5.9) for patients with low BMI and for patients with normal or higher BMI, respectively. In the multivariate analysis, low BMI (95% CI 1.07–4.44), ECOG ≥2 (95% CI 1.02–2.89), hypoalbuminemia (95% CI 1.21–4.01), high LDH (95% CI 1.22–3.49), and light chain ratio > 100 (95% CI 1.06–2.77) were independent risk factors of mortality. MM patients who were underweight, with hypoalbuminemia, poor performance status, higher LDH, and light chain ratio > 100 were associated with poor overall survival.

Depending on the scanning mode, existing short text stream clustering methods can be divided into the following two kinds of methods: one-pass-based and batch-based. The one-pass-based method handles each text only one time, but cannot deal with the sparseness problem very well. The batch-based method obtains better results by allowing multiple iterations of each batch, but the efficiency is relatively low. To overcome these problems, this paper presents Lifelong learning Augmented Short Text stream clustering method (LAST), which incorporates the episodic memory module and sparse experience replay module of lifelong learning into the clustering process. Specifically, LAST processes each text one time, but at a certain interval it randomly samples some previously seen texts of the episodic memory to update cluster features by performing sparse experience replay. Empirical studies on two public datasets demonstrate that the performance of the LAST-based method is on a par with the batch-based method, and runs close to the speed of the one-pass-based method.

Multi-label classification aims to deal with the problem that an object may be associated with one or more labels, which is a more difficult task due to the complex nature of multi-label data. The crucial problem of multi-label classification is the more robust and higher-level feature representation learning, which can reduce non-helpful feature attributes from the input space prior to training. In recent years, deep learning methods based on autoencoders have achieved excellent performance in multi-label classification for the advantages of powerful representations learning ability and fast convergence speed. However, most existing autoencoder-based methods only rely on the single autoencoder model, which pose challenges for multi-label feature representations learning and fail to measure similarities between data spaces. To address this problem, in this paper, we propose a novel representation learning method with dual autoencoder for multi-label classification. Compared to the existing autoencoder-based methods, our proposed method can capture different characteristics and more abstract features from data by the serially connection of two different types of autoencoders. More specifically, firstly, the algorithm of Reconstruction Independent Component Analysis (RICA) in sparse autoencoder is trained on patches on all training and test dataset for robust global feature representations learning. Secondly, with the output of RICA, stacked autoencoder with manifold regularization (SAMR) is introduced to ameliorate the quality of multi-label features learning. Comprehensive experiments on several real-world data sets demonstrate the effectiveness of our proposed approach compared with several competing state-of-the-art methods.

Lexical simplification has attracted much attention in many languages, which is the process of replacing complex words in a given sentence with simpler alternatives of equivalent meaning. Although the richness of vocabulary in Chinese makes the text very difficult to read for children and non-native speakers, there is no research work for the Chinese lexical simplification (CLS) task. To circumvent difficulties in acquiring annotations, we manually create the first benchmark dataset for CLS, which can be used for evaluating the lexical simplification systems automatically. To acquire a more thorough comparison, we present five different types of methods as baselines to generate substitute candidates for the complex word that includes synonym-based approach, word embedding-based approach, BERT-based approach, sememe-based approach, and a hybrid approach. Finally, we design the experimental evaluation of these baselines and discuss their advantages and disadvantages. To our best knowledge, this is the first study for CLS task.

Anlotinib, a highly selective multi-targeted tyrosine kinase inhibitor (TKI) has therapeutic effects on non-small-cell lung cancer (NSCLC). In this study, the anti-tumor activity and molecular mechanism of anlotinib in metastatic colorectal cancer (mCRC) was explored. The anti-angiogenesis, anti-metastasis, anti-proliferative, and anti-multidrug resistance efficacy of anlotinib were analyzed by using in vitro and in vivo models of human CRC cells. The results indicated that anlotinib boosted chemo-sensitivity of CRC cells, and restrained its proliferation. Besides the suppression of the MET signaling pathway, anlotinib also inhibited invasion and migration of CRC cells. Furthermore, anlotinib prevented VEGF-induced angiogenesis, N-cadherin (CDH2)-induced cell migration, and reversed ATP-binding cassette subfamily B member 1 (ABCB1) -mediated CRC multidrug resistance in CRC. The CRC liver metastasis and subcutaneously implanted xenograft model testified that anlotinib could inhibit proliferation and liver metastasis in CRC cells. Such an observation suggested that a combination of anlotinib with anti-cancer drugs could attenuate angiogenesis, metastasis, proliferative, and multidrug resistance, which constitutes a novel treatment strategy for CRC patients with metastasis.

Deep hashing has greatly improved retrieval performance with the powerful learning capability of deep neural network. However, deep unsupervised hashing can hardly achieve impressive performance due to the lack of the semantic supervision. This letter proposes Unsupervised Discriminative Deep Hashing (UD $^2$ H) to fulfill this gap. UD $^2$ H is formulated to jointly perform hash code learning and clustering, and trained in an asymmetric manner to improve the efficiency. The cluster labels supervise the training of deep model to enable hash code discriminative. Based on the outputs of the deep model, UD $^2$ H adaptively constructs a similarity graph that considers the local and global structures. Experiments on three benchmark datasets show that the proposed UD $^2$ H outperforms the state-of-the-art unsupervised deep hashing methods.

Lexical simplification (LS) aims at replacing complex words with simpler alternatives. LS commonly consists of three main steps: complex word identification, substitute generation, and substitute ranking. Existing LS methods focus on the contextual information of the complex word in the last step (substitute ranking). However, they miss out the following two facts: (1) The word complexity of a polysemous word is very closely related to its context; (2) The step of substitute generation regardless of the context will inevitably produce a large number of spurious candidates. Therefore, we propose a novel LS system LSBert based on pretrained language model BERT to address the aforementioned issues, which is capable of making use of the wider context when both identifying the words in need of simplification and generating substitute candidates for the complex words.Specifically, LSBert consists of a network for complex word identification by fine-tuning BERT and a network for substitute generation based on BERT. Experimental results show that LSBert performs well in both complex word identification and substitute generation, achieving state-of-the-art results in three benchmarks. To facilitate reproducibility, the code of the LSBert system is available at https://github.com/qiang2100/BERT-LS .

To evaluate the prevalence of Helicobacter pylori infection and risk factors and to serotype the strains in Wuwei, located in north-western China, which has a high incidence of gastric cancer.Helicobacter pylori infection was analysed in 21 291 adults byThe infection rate of H. pylori was 53.0%, and 90.1% of strains were type I strains. The H. pylori infection rate was higher among farmers (OR = 1.34, 95% CI: 1.19-1.50) and individuals who had a junior high school or higher education level (OR = 1.10, 95% CI: 1.06-1.15), and was lower in older individuals (OR = 0.86, 95% CI: 0.83-0.90), individuals with high income (OR = 0.93, 95% CI: 0.90-0.95), individuals with a habit of eating quickly (OR = 0.93, 95% CI: 0.87-0.99) and individuals who consumed more fruit and vegetables (OR = 0.90, 95% CI: 0.85-0.95). Individuals with history of cholecystitis/cholecystolithiasis, hypertension and asthma were negatively correlated with H. pylori infection (P 0.05).The prevalence of H. pylori infection is high in Wuwei. The major prevalent strain is type I strain. Age, education, occupation, household income, consumption of fruit and vegetables, and habit of eating quickly are independent risk factors for H. pylori infection, which is also associated with individuals with a history of extragastric diseases.Evaluer la prévalence de l'infection à Helicobacter pylori et les facteurs de risque et déterminer le sérotype des souches à Wuwei, situé dans le nord-ouest de la Chine, où l'incidence du cancer gastrique est élevée. MÉTHODES: L'infection à H. pylori a été analysée chez 21.291 adultes par un test respiratoire à l'urée auLa prévalence de l'infection à H. pylori est élevée à Wuwei. La principale souche répandue est du type I. L'âge, l'éducation, la profession, le revenu du ménage, la consommation de fruits et de légumes et l'habitude de manger rapidement sont des facteurs de risque indépendants d'infection à H. pylori, qui est également associée à des personnes ayant des antécédents de maladies extra-gastriques.

BACKGROUND: Noise is a hazard for newborns. Preterm infants are more vulnerable to negative effects of noise because their auditory system is at a critical period of neurodevelopment. We conducted this study to determine whether noise exposure in our Level II unit met the American Academy of Pediatrics (AAP)-issued recommendation of less than 45 dB. We also aimed to assess the efficacy of a bundle of noise reduction methods and the role of noise supervisors in assuring the compliance of modification. METHODS: Hourly mean equivalent continuous noise level (Leq) was collected in two phases: (I) Phase 1: baseline data for 4 weeks, (II) Phase 2: study period for 13 weeks, 1 week (Stage 1) under close monitoring by the noise supervisors and 12 weeks (Stages 2, 3 and 4, lasting 4 weeks each) without supervisors. RESULTS: The baseline noise level consistently exceeded recommendations with an hourly mean incubator noise of 53.6 dB (±5.2). Our bundle resulted in a significant reduction in incubator noise levels by 9.1 dB (±0.75). The sound levels remained

With the rapid development of optoelectronics, methods for manipulating light–matter interactions are urgently required. In this work, three-dimensional double split-ring resonators (3D-DSRRs) have...

Rail surface defect detection plays a critical role in the maintenance of the rail transportation system. Video analysis technology is a promising method to detect defects due to its low cost and effectiveness. Several attempts with hand-craft features have been made to obtain the detection results by using traditional machine vision algorithms. However, these methods suffer from imprecise results due to challenging conditions, such as deteriorated and changeable lighting environment and various types of complex rail surface defects. Recently, classification methods with complex deep convolutional networks have become popular. Despite their high accuracy, these methods cannot meet the requirements of defects localisation and real-time processing in practice. To solve these problems, this study proposes a novel object detection algorithm to detect rail defects. The net architecture of the proposed algorithm includes a backbone network using MobileNet and several novel detection layers with multi-scale feature maps inspired by you only look once (YOLO) and feature pyramid networks. Two different architectures of MobileNet are used to estimate the performance of defects detection. The experimental results demonstrate the great potential of the proposed algorithm with fast inference speed and high accuracy in the industry.

During underground mining, accurate revelation on the deformation and failure mechanisms of a high-steep slope under multi-layer mining conditions facilitates the prevention and control of geological disasters in mines. Numerical simulation based on discrete element theory can be used to explore the characteristics and mechanism of action of deformation and failure of a slope under complex geological and multi-layer mining conditions. By utilising PFC2D (particle flow code) software, the deformation and failure characteristics of a high-steep slope in Faer Coal Mine in Guizhou Province, China were investigated. Additionally, the mechanism of influence of different numbers of mining layers on the deformation and failure of the high and steep slope was elucidated. The result showed that after the goaf passed by the slope toe, multi-layer mining aggravated the subsidence and deformation of the slope toe: the slope toppled forward as it sank. The toppling of the slope changed the slope structures: the strata in the front of the slope were transformed from anti-dip to down-dip features. Extruded by collapsed–toppled rock mass, the slope toe and the rock mass located in the lower part of the slope toe generally exhibited a locking effect on the slope. Multi-layer mining degraded the overall stability of the slope, in that the total displacement of the slope was much greater than the total mining thickness of the coal seams. Based on the aforementioned research, ideas for preventing and controlling geological disasters during mining operations under a high-steep slope were proposed.

Nickel oxide (NiOx) is a promising hole-transporting material for perovskite light-emitting diodes (PeLEDs) because of its low cost, excellent stability, and simple fabrication process. However, th...

Information technology and advanced online environments have reduced the cost of these exchange activities and triggered the emergence of the sharing economy. Con-sequently, public attitude toward the sharing economy has gradually shifted from re-luctance to acceptance. Moreover, the sharing economy has revolutionized the busi-ness models and viewpoints of conventional industries, and sharing service providers have gradually shifted from an independent to a collaborative stance, thereby affect-ing conventional economies. This study interprets the phenomenon of cross-industry collaboration in the sharing economy through social exchange and social network the-ories. A multiple-case research framework is used to examine tourism and service in-dustries. Secondary data of service providers and users on sharing platforms are ana-lyzed using content analysis, supplemented with a content analysis of the interview data of three hotel executives. The varying phenomena of the conventional and shar-ing economies on social exchange and social network were compared. Finally, this paper proposes conclusions and practical recommendations according to the analytical results. JEL classification numbers: D85, M31, L14. Keywords: Cross-Industry Collaboration, Sharing Economy, Social Exchange, Social Network.

Forecasting the crowd flows of some densely populated areas (functional areas) at urban level is vital for the city and traffic management and much efforts have been put into this field. However, considering the complex dependencies of crowd flows prediction, the current models generally neglected the multi-scale correlation of spatial view and weakened the influence of external features on crowd flows data. In this paper, we proposed a deep learning framework, called MV-RANet, to deal with them and further improve the prediction performance. Firstly, a double-branch residual attention network was utilized to model the spatial correlation between fine-grained regions and coarse-grained functional areas separately with different branch. And furthermore, some human mobility pattern information were incorporated in MV-RANet to capture the correlation between functional areas in a different view. And as for the impact of external factors, we combined them with crowd flows data by multi-channel mechanism, which can make the external factors and spatio-temporal crowd flows data one-to-one mapping, to refine their influence on the crowd flows. We evaluated our model with two different datasets, a large scale Call Detail Records (CDRs) data and Beijing taxicabs’ trajectories data (TaxiBJ) (Zhang et al., 2017). The experimental results demonstrate that the proposed MV-RANet consistently outperforms all the baselines.

Ultrafast fiber lasers with a wide spectral range have huge demands for many applications. In this article, magnetron sputtering deposition (MSD) technique was utilized to prepare a new broadband microfiber based Sb thin film saturable absorber (SA), and ultrafast pulses in both 1.5 and 2 μm wavebands were generated with the integrated microfiber-based Sb thin film SA in erbium-doped fiber laser (EDFL) and thulium-doped fiber laser (TDFL), respectively. Pulse duration of 544 fs/972 fs and output power of 21.76 mW/36.1 mW centered at 1557 nm/1892 nm were respectively produced from stable mode-locked EDFL and TDFL. To the best our knowledge, this is the first time that ultrafast fiber lasers including both 1.5 μm and 2 μm wavebands were mode locked by a single microfiber-based Sb thin film SA.

Attosecond light source is a new type of light source that was born at the beginning of the 21st century, which has a short pulse, broad spectrum, high temporal and spatial coherence and wide tunability, thus being widely employed in various research fieds. From ultrafast motion of electrons in atoms to charge transfer in biological macromolecules, attosecond pulses is currently the only tool that can track and capture these ultrafast dynamics. Attosecond pulse enables us to investigate ultrafast dynamics of micro-world both in nanometer and attosecond scales. However, the mechanism of attosecond pulse generation is completely different from general ultrafast lasers. Instead, attosecond pulse is generated by a highly nonlinear interaction between strong ultrafast femtosecond laser and matter, which is called high-order harmonic generation (HHG). The mechanism of HHG can be understood by a classical three step model. First, an electron is ionized from an atom by a strong laser electric field through tunnel ionization. Then the free electron is accelerated by the laser field and gains energy. Finally, the electron recombines with the parent ion when the laser field changes its sign with emission of a photon, whose energy equals its kinetic energy gained in the laser field plus ionization potential of the atom. Apparently, HHG is strongly affected by the laser waveform. Several key parameters of driving laser, such as wavelength, intensity, and carrier envelope phase strongly affect the process of HHG. Thus, the characteristics of attosecond pulses are determined by the driving laser. The rapid development of attosecond pulse technology strongly depends on the development of driving laser technology. In the beginning, chirped pulse amplification (CPA) technology greatly promoted the development of attosecond light sources. The femtosecond CPA systems based on Ti:sapphire crystal has been the main driving laser to generate attosecond light pulses. The driving laser wavelength is in the near infrared region near 800 nm, which generally has a pulse duration of multiple optical cycles. By employing post-compression technology to shorten pulse durations of CPA systems to few-cycle, isolated attosecond pulses in the extreme ultraviolet (XUV) can be generated, which is called the first generation of attosecond light source. Recently, optical parametric amplification (OPA) systems have been widely used as driving laser due to its flexible wavelength tunability. Using OPA, longer driving laser wavelength up to the midinfrared (MIR) can be obtained, which pushes the attosecond pulses to the soft X-ray region, and has been called the second generation of attosecond light sources. Due to broad spectrum and higher photon energy, attosecond pulses have a shorter duration in the soft X-ray region compared with XUV region, given that the temporal chirp is properly compensated. In 2017, reseachers generated soft X-ray isolated attosecond pulses, which was driven by mid-infrared pulses centered at 1.8 µm. These attosecond pulses, whose duration reaches 53 as and spectrum surpasses carbon K-edge, provide a tool for studying the ultrafast dynamics of diamond, graphene and other carbon materials. In this paper, we start with the principle of attosecond pulses generation based on HHG. Then, we introduce different technologies and their development for obtaining driving laser pulses for HHG. Finally, we introduce prospect on development of driving laser pulses for attosecond pulses generation.

Infection is associated with great morbidity and mortality in patients with multiple myeloma (MM), but evidence for invasive fungal infections (IFIs) is lacking. We aimed to investigate risk factors for IFI in MM patients and to determine its impact on patients’ survival. We retrospectively analyzed MM patients at Taipei Veterans General Hospital in Taiwan between January 2002 and October 2018. MM was diagnosed according to the International Myeloma Working Group criteria. IFI was defined according to the European Organization for Research and Treatment of Cancer/Mycoses Study Group criteria. All risk factors of IFI in MM patients were estimated using Cox regression models in the univariate and multivariate analyses. Of the 623 patients recruited, 22 (3.5%) were diagnosed with proven or probable IFI. Light chain disease (adjusted hazard ratio [HR] 6.74, 95% confidence interval [CI] 2.10–21.66), hemoglobin less than 8 g/dl (adjusted HR 3.34, 95% CI 1.32–8.42), serum albumin

Graph embedding has attracted much more research interests in dimensionality reduction. In this study, based on collaborative representation and graph embedding, the authors propose a new linear dimensionality reduction method called collaborative representation-based locality preserving projection (CRLPP). In the CRLPP, they assume that the similar samples should have similar reconstructions by collaborative representation and the similar reconstructions should also have the similar low-dimensional representations in the projected subspace. CRLPP first reconstructs each training sample using the collaborative representation of the other remaining training samples, and then designs the graph construction of all training samples, finally establishes the objective function of graph embedding using the collaborative reconstructions and the constructed graph. The proposed CRLPP can well preserve the intrinsic geometrical and discriminant structures of high-dimensional data in low-dimensional subspace. The effectiveness of the proposed is verified on several image datasets. The experimental results show that the proposed method outperforms the state-of-art dimensionality reduction.

Visualizing the details of different cellular structures is of great importance to elucidate cellular functions. However, it is challenging to obtain high quality images of different structures directly due to complex cellular environments. Fluorescence staining is a popular technique to label different structures but has several drawbacks. In particular, label staining is time consuming and may affect cell morphology, and simultaneous labels are inherently limited. This raises the need of building computational models to learn relationships between unlabeled microscopy images and labeled fluorescence images, and to infer fluorescence labels of other microscopy images excluding the physical staining process. We propose to develop a novel deep model for virtual staining of unlabeled microscopy images. We first propose a novel network layer, known as the global pixel transformer layer, that fuses global information from inputs effectively. The proposed global pixel transformer layer can generate outputs with arbitrary dimensions, and can be employed for all the regular, down-sampling, and up-sampling operators. We then incorporate our proposed global pixel transformer layers and dense blocks to build an U-Net like network. We believe such a design can promote feature reusing between layers. In addition, we propose a multi-scale input strategy to encourage networks to capture features at different scales. We conduct evaluations across various fluorescence image prediction tasks to demonstrate the effectiveness of our approach. Both quantitative and qualitative results show that our method outperforms the state-of-the-art approach significantly. It is also shown that our proposed global pixel transformer layer is useful to improve the fluorescence image prediction results., Comment: 10 pages, 6 figures, 5 tables

Line-commutated converter based high voltage direct current (LCC-HVDC) transmissions are widely applied in delivering large-capacity electricity in China. Multiple large-capacity LCC-HVDC transmissions strengthen dynamic coupling between AC and DC networks. Existing literature has reported some stability problems caused by dynamic interactions between AC and DC power grids, however, the mechanism behind these problems has not been totally understood and revealed. To analyze dynamic interactions between AC and DC power grids, this article presents a small-signal model to understand the external characteristics of LCC-HVDC stations based on the motion equation concept. Comparison of time-domain simulation response with the detailed switching model shows that the proposed model can hold main behaviors of concern. Dynamic interactions between AC and DC power grids are studied based on the proposed model in multi-infeed LCC-HVDC systems. Analytical results demonstrate that there exists power dynamics in receiving AC grid when sending AC grid subjects to a disturbance, and effects of converter stations’ controller parameters on power dynamics are explained by the physical relationship among state variables of the proposed model.

BACKGROUND Thrombotic microangiopathy (TMA) syndromes are potentially life-threatening complications and are defined as integrated syndromes of microangiopathic hemolytic anemia, thrombocytopenia, and organ injury. Systemic lupus erythematosus (SLE) is an autoimmune disease that can affect various organs, including the hematopoietic system. SLE can complicate with TMA and can be categorized into two distinct groups by chronological association: TMA occurring as the initial presentation and leading to a diagnosis of SLE concurrently (TMA-cSLE) or TMA developing in patients previously diagnosed as having SLE (TMA-pSLE). We examined the differences in clinical characteristics, treatment responses, and clinical outcomes between these groups. METHODS We reviewed data of patients diagnosed as having TMA and SLE at Taipei Veterans General Hospital between 2002 and 2013. We included 29 patients: 8 and 21 in TMA-cSLE and TMA-pSLE groups, respectively. All underwent plasma exchange. Patients' demographic and clinical characteristics, disease activity, and treatment modality were summarized. RESULTS Overall survival (OS) from SLE or TMA diagnosis was poor for the TMA-cSLE group. Median OS from SLE diagnosis was 2.9 months in the TMA-cSLE group and 103.5 months in the TMA-pSLE group (p < 0.001). Median OS from TMA diagnosis was 2.9 months in the TMA-cSLE group and 10.7 months in the TMA-pSLE group (p = 0.58). Time to TMA remission after treatment appeared longer in the TMA-cSLE group (38.00 vs 10.76 days). Multivariate Cox analysis revealed TMA-cSLE and anti-RNP positivity were independent risk factors for mortality in SLE patients with TMA. CONCLUSION The occurrence of TMA with SLE is rare, and its vigorous course results in high mortality and morbidity rates. In patients without a history of autoimmune disease, early suspicion of TMA and working-up for SLE under this condition are vital. Early recognition of TMA-cSLE and prompt plasma exchange with upfront immunosuppressive therapies for TMA-cSLE patients or anti-RNP-positive patients may improve their prognosis.

Introduction: Klebsiella pneumoniae, a common hospital- and community-acquired pathogen, is notorious for multidrug resistance. This study aimed to better understand the correlation of clinical presentation and microbiological characteristics of the isolates causing bloodstream infections (BSIs) in Taiwan. Methodology: We retrospectively collected 150 isolates derived from K. pneumoniae bacteremia patients in Taiwan in both 2014 and 2016. Clinical data, bacterial serotyping and drug susceptibility tests were comparatively analyzed. Results: Demographic data showed that diabetes mellitus (DM) was the most common underlying disease (44.0%). The overall 30-day mortality rate was 19.3%, and higher mortality was found in patients with malignancy than others (P = 0.023). Serotype distribution was diverse. The major isolates belonged to non-PCR-typeable serotypes (58.7%) associated with hospital-acquired infections (P = 0.007) and in non-DM patients (P < 0.001), while K2 and K20 significantly caused infections and in DM patients (P = 0.046 and P = 0.006, respectively); however, only K2 showed more community-acquired infection (P = 0.022) than other typeable serotypes. Resistance to antibiotics in clinical isolates in the year 2016 was > 24%, including cefazolin (54%), ampicillin-sulbactam (25%) and cefuroxime (25%). Susceptibility to gentamicin, flomoxef, and tigecycline reduced between the two time periods (2014 and 2016). However, the isolates remained highly susceptible to amikacin and ertapenem (> 95%). Conclusions: Patients with cancer had a higher 30-day mortality rate than others. Amikacin and ertapenem are the drugs of choice for the treatment of multidrug-resistant K. pneumoniae BSIs in Taiwan.