Search

Two-dimensional monolayer and few-layer transition-metal dichalcogenides (TMDs) are promising for advanced electronic and photonic applications due to their extraordinary optoelectronic and mechanical properties. However, it has remained challenging to produce high-quality TMD thin films with controlled thickness and desired micropatterns, which are essential for their practical implementation in functional devices. In this work, a self-limiting opto-electrochemical thinning (sOET) technique is developed for on-demand thinning and patterning of TMD flakes at high efficiency. Benefiting from optically enhanced electrochemical reactions, sOET features a low operational optical power density of down to 70 μW μm-2 to avoid photodamage and thermal damage to the thinned TMD flakes. Through selective optical excitation with different laser wavelengths based on the thickness-dependent band gaps of TMD materials, sOET enables precise control over the final thickness of TMD flakes. With the capability of thickness control and site-specific patterning, our sOET offers an effective route to fabricating high-quality TMD materials for a broad range of applications in nanoelectronics, nanomechanics, and nanophotonics.

Aiming at the characteristics of low intelligence of the current motor cluster and superconducting electrical equipment monitoring system, high construction and maintenance costs of wired transmission methods. The wireless sensor network constructed by LoRa technology is applied to the motor operating state monitoring system, while the data transmission efficiency of the system is improved by optimizing the topology of the wireless sensor network (WSN) and processing the transmission data redundancy. The system uses STM32F407ZET6 as the main controller, builds an offline operating state library, and uses HMM model for training to achieve real-time acquisition and fault diagnosis of the status information of the cluster system motor's voltage, current, speed, and position. Practical application shows that the system is running well and can realize the acquisition of the motor's running status and fault diagnosis.

In order to realize the research on network redundant data cleaning based on big data, this paper designs a set of redundant data cleaning framework according to the data processing flow before data analysis. According to the spatial correlation of redundant data, a method of data cleaning is designed. In the data cleaning method, appropriate cleaning algorithms are designed for abnormal data and missing data respectively, in which mathematical probability design is applied to abnormal data to delete the data with obvious deviation from the normal data value. The spatial model and algorithm are designed by applying spatial correlation to the missing data to fill the missing data value after the redundant data is cleaned by other steps in the method. The accuracy of the model is compared with that of the common data prediction algorithm, and the accuracy between the algorithm and the redundant data set is verified.

Purpose: Magnesium-based alloy scaffold is a promising biodegradable stent due to its intrinsic mechanical performance and biocompatibility. Based on our preliminary experiments, we designed a novel sirolimus-eluting magnesium-based alloy scaffold. This work aimed to assess its safety and degradation performance in vivo. Methods: The scaffolds were implanted in the lower limb arteries of Bama mini-pigs. Safety was defined as no immediate thrombosis or >30% residual stenosis, which was assessed with optical coherence tomography and digital subtraction angiography. Blood biochemical analyses were performed to evaluate hepatorenal toxicity. The degradation process of the scaffolds, the endothelialization, and lumen loss of the stented-vessels were detected with scanning electron microscopy, immunohistochemical, hematoxylin-eosin staining and optical coherence tomography. Results: Twenty-four scaffolds were successfully implanted in six pigs with no signs of immediate thrombosis or >30% residual stenosis. The scaffolds were covered by endothelium at one month and absolutely resorbed at six months post implantation. Blood analysis showed that the hepatorenal function except for alanine aminotransferase and γ-glutamyl transpeptidase was normal. Obvious intimal hyperplasia and lumen loss were found in the stented vessels at three months, while the diameters and inner lumen areas of stented segments had increased significantly at six months (p

Purpose To repair multiple skin soft tissue defects of one finger is a challenge to hand surgeons. We introduce a method which can be used to repair multiple skin soft tissue defects of one finger with bilateral flaps in parallel pattern flap based on the dorsal branches of the proper digital artery. Method A patient suffered electric injury in her left index finger with two soft tissue defects, and the areas were 1.6 cm × 1.0 cm and 2.2 cm × 1.2 cm, respectively. And who underwent a homodigital parallel flaps based on the dorsal branches of the proper digital artery to repair in January 2018. The donor sites were covered by full-thickness skin grafting. Results The flaps and the skin grafting survived uneventfully. All incisions achieved primary healing. The follow-up was 19 months, and the shape of the flaps was satisfactory with soft texture and suitable appearance. TAM of the injured finger was 210°, the level was excellent. The score of VAS was 9. Conclusion The homodigital bilateral flaps in parallel pattern based on the dorsal branches of the proper digital artery are a potential treatment in one-stage for multiple skin soft tissue defects of one finger with reliable blood supply, satisfactory results and simple surgical procedure.

Recent developments in deep neural network (DNN) pruning introduces data sparsity to enable deep learning applications to run more efficiently on resource- and energy-constrained hardware platforms. However, these sparse models require specialized hardware structures to exploit the sparsity for storage, latency, and efficiency improvements to the full extent. In this work, we present the sparse neural acceleration processor (SNAP) to exploit unstructured sparsity in DNNs. SNAP uses parallel associative search to discover valid weight (W) and input activation (IA) pairs from compressed, unstructured, sparse W and IA data arrays. The associative search allows SNAP to maintain a 75% average compute utilization. SNAP follows a channel-first dataflow and uses a two-level partial sum (psum) reduction dataflow to eliminate access contention at the output buffer and cut the psum writeback traffic by 22 $\times $ compared with state-of-the-art DNN accelerator designs. SNAP’s psum reduction dataflow can be configured in two modes to support general convolution (CONV) layers, pointwise CONV, and fully connected layers. A prototype SNAP chip is implemented in a 16-nm CMOS technology. The 2.3-mm2 test chip is measured to achieve a peak effectual efficiency of 21.55 TOPS/W (16 b) at 0.55 V and 260 MHz for CONV layers with 10% weight and activation densities. Operating on a pruned ResNet-50 network, the test chip achieves a peak throughput of 90.98 frames/s at 0.80 V and 480 MHz, dissipating 348 mW.

Visual-based road crack detection becomes a hot research topic over the last decade because of its huge application demands. Road crack detection is actually a special form of salient object detection task, whose objects are small and distribute randomly in the image compared to the traditional ones, which increase the difficulty of detecting. Most conventional methods utilize bottom information, such as color, texture, and contrast, to extract the crack regions in the image. Even though these methods can achieve satisfactory performances for images with simple scenarios, they are easily interfered by some factors such as light and shadow, which may decrease the detection result directly. Inspired by the competitive performances of deep convolutional neural networks on many visual tasks, we propose a distribution equalization learning mechanism for road crack detection in this paper. Firstly, we consider the crack detection task as a pixel-level classification and use a U-Net based architecture to finalize it. Secondly, the occurrence probability of crack and non-crack are so different, which results in the ill-conditioned classifier and undesirable detection performance, especially the high false detection rate. In this case, we propose a weighted cross entropy loss term and a data augmentation strategy to avoid influence from imbalanced samples through emphasizing the crack regions. Additionally, we propose an auxiliary interaction loss, and combine it with the popular self-attention strategy to alleviate the fracture situations through considering relationships among different local regions in the image. Finally, we tested the proposed method on three public and challenging datasets, and the experimental results demonstrate its effectiveness.

The microstructure of the ceramic topcoat has a great influence on the service performance of thermal barrier coatings (TBCs). In this study, conventional layered-structure TBCs, nanostructured TBCs, and novel-structured TBCs with a unique microstructure were fabricated by air plasma spraying. The relationship between the microstructure and properties of the three different TBCs was analysed. Their thermal insulation ability, sintering resistance, and durability were systematically evaluated. Additionally, their failure modes after being subjected to two kinds of thermal shock tests were analysed. The results revealed that the novel-structured TBCs had remarkably superior performances in all the examined aspects. The thermal conductivity of the novel-structured TBCs was significantly lower than those of the conventional and nanostructured TBCs both in the as-sprayed state and after thermal treatment for 500 h at 1100 °C. The macroscopic elastic modulus of the novel-structured TBCs after sintering at 1300 °C for 100 h was similar to those of the conventional and nanostructured TBCs in the as-sprayed state. During both a burner rig thermal shock test and a furnace cyclic oxidation test, the thermal shock lifetime of the novel-structured TBCs was much longer than those of the conventional and nanostructured TBCs. This study has demonstrated novel-structured plasma-sprayed TBCs with high thermal insulation ability and high durability.

Reusable rocket engines are the core components of reusable launch vehicles, and have thus become a major focus of aerospace engineering research in recent years. In practice, subsystem design is based on the overall index allocation of an engine; therefore, a multidisciplinary optimization approach is necessary. In this study, design of a reusable methane/liquid oxygen (LOX/CH4) rocket engine with a gas generator cycle was investigated using multidisciplinary optimization. Two parameters were chosen as design variables: pressure and fuel mix ratio of the main combustion chamber. Optimization objectives were specific impulse, structural mass, and life cycle cost of the reusable rocket engine, and constraints were assigned to each discipline according to rocket design requirements. Then, an optimization model was developed, and optimal design parameters were acquired for the LOX/CH4 rocket engine. The proposed method is effective for designing the index allocation of reusable rocket engines and takes into account the multidisciplinary nature of complex systems.

Objective Various modalities are applied for pathological diagnosis of malignant biliary strictures (MBS), including brush cytology (BC), forceps biopsy (FB) and endoscopic ultrasound-guided fine needle aspiration (EUS-FNA). We aimed to assess the value of these modalities in a repeated tissue acquisition process for biliary strictures with initially inconclusive pathological outcomes. Methods Patients who were suspected of having MBS and underwent a BC in two large teaching hospitals were retrospectively included. The sensitivity, specificity, positive and negative predictive values, and accuracy of the initial and repeated BC, FB and EUS-FNA were analyzed. Their performances were compared to determine which modality was superior in repeated tissue acquisition. Results In total, 476 patients were included. The sensitivity, specificity and accuracy in diagnosing MBS for the initial BC were 30.3%, 100% and 55.0%, respectively. Altogether 39, 27 and 44 patients underwent a repeat BC, FB and EUS-FNA, respectively. The sensitivity for repeated BC, FB and EUS-FNA was 41.2%, 61.1% and 44.4%, respectively, whereas their specificity all reached 100%. When comparing diagnostic accuracy, none of the modalities was superior (74.4% vs 74.1% vs 54.5%, P = 0.173). In the repeated process, one patient who underwent BC and two underwent FB developed mild pancreatitis. Conclusions Repeated tissue acquisition achieves a conclusive diagnosis of MBS in nearly half patients who have an initially inconclusive cytological diagnosis. None of the tissue acquisition methods is significantly superior in the repeated process.

Global positioning system (GPS) trajectory map matching projects GPS coordinates to the road network. Most existing algorithms focus on the geometric and topological relationships of the road network, while did not make full use of the historical road network information and floating car data. In this study, the authors proposed a deep learning enabled vehicle trajectory map-matching method with advanced spatial–temporal analysis (DST-MM). The algorithm mainly focused on the following three aspects: (i) analyse the spatial relevancy from the prospective of geometric analysis, topology analysis and intersection analysis; (ii) to make full use of the historical and real-time data, a deep learning model was conducted to extract the road network and vehicle trajectory features and (iii) establish a speed prediction model and nest it in the temporal analysis structure. It narrows down the path search range through establishing the dynamic candidate graph. Experimental results show that the proposed DST-MM algorithm outperforms the existing algorithms in terms of matching accuracy for low-sampling frequencies GPS data, especially in the central urban area.

Introduction High altitude polycythemia (HAPC) is a common chronic disease at high altitudes. It is characterized by excessive erythrocytosis (≥190 g·L-1 in females or ≥210 g·L-1 in males). HAPC severely jeopardizes the health status of plateau dwellers. The Qinghai-Tibet plateau, with an elevation above 4000 m, is the highest plateau in the world. Both Han and Tibetan populations residing there face the threat of HAPC. Therapeutic erythrocytapheresis (TE) was introduced to Tibet as an alternative to phlebotomy in 2015. Methods In this study, we retrospectively analyzed 155 patients with HAPC treated with TE in Tibet. Routine blood testing values before and after TE were compared to evaluate treatment efficacy. The efficiency rate, defined as the rate of increase in red blood cell depletion attained by TE compared with 450 mL whole blood phlebotomy, was calculated using whole blood volume and hematocrit before and after treatment and used to identify patients who maintained a normal hemoglobin level in the year after the TE procedure. Results On average, TE reduced red blood cell levels by 1.5×1012·L-1, hemoglobin concentration by 52 g·L-1, and hematocrit by 14% (P Conclusions TE is a feasible therapeutic method to treat HAPC on the Qinghai-Tibet plateau. The efficiency rate is a useful tool to predict the expected interval between TE procedures.

Background: The objective was to evaluate the effects of personal characteristics on the validation of self-reported type 2 diabetes among Chinese adults in urban Shanghai. Methods: During 2015 through 2016, 4,322 participants were recruited in this validation study. We considered the criteria of diabetes verification to use the laboratory assays of fasting plasma glucose (FPG), glycated hemoglobin (HbA1c), or self-reported use of diabetic medication. Results: When taking diabetic medication or FPG ≥7.0 mmol/L was as identified diabetes, the measurements of sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and Kappa value of self-reported diabetes were 72.0%, 99.2%, 95.1%, 93.9%, and 0.78, respectively. If an additional HbA1c test was used for 708 subjects (aged

Background Tumor cells undergoing epithelial-mesenchymal transition (EMT) display enhanced ability to enter the circulation, thereby being major source of circulating tumor cells (CTCs). In this study, we aimed to better understand the roles of CTC undergoing EMT in monitoring cancer progression. Methods We analyzed gene expression profiling of epithelial and mesenchymal markers in lung or colon tumor samples by mining TCGA database. We detected CTCs and classify their EMT phenotypes of 31 patients with lung or colon cancer by using a CanPatrol CTC-enrichment technique. Results The bioinformatic analysis indicated that mesenchymal markers were expressed in a subset of lung tumor samples, and its high expression was associated with poor survival of lung cancer patients. However, in colon cancer, majority of tumor samples expressed hybrid epithelial/mesenchymal markers. CTC analysis with EMT classification showed that the number of CTCs with mesenchymal phenotype was high in lung cancer patients with the advanced stage. Dynamic CTC analysis in a lung cancer patient indicated that CTC with mesenchymal phenotype was effective to monitor tumor progression. In a colon cancer patient, dynamic CTC analysis indicated that CTC with hybrid epithelial/mesenchymal phenotypes was an effective biomarker to guide therapy. Conclusions Encouraging results from this proof-of-concept study show that CTC with mesenchymal phenotype or hybrid epithelial/mesenchymal phenotypes could be a potential biomarker for monitoring tumor progression in lung or colon cancer respectively.

Instead of conducting manual factor construction based on traditional and behavioural finance analysis, academic researchers and quantitative investment managers have leveraged Genetic Programming ...

Classification is a fundamental technique in hyperspectral image understanding field, which is to give each pixel a specific semantic label based on its characteristics automatically. Recently, even though convolutional neural network (CNN) based methods have been applied to hyperspectral image classification, their performances are not in accordance with our expectation. The reason is that, most of the existing methods can not exploit the intrinsic characteristics of different pixels in hyperspectral image effectively, they often ignore the inherent relationships among different spatial pixels, different spectral bands and different category dependencies. Specifically, the relationships among different spatial pixels mean that the similarities of attribute and location information among different pixels. The relationships among different spectral bands mean that the different contributions of them to the final classification. The relationships among different categories mainly mean that the severe sample imbalance problem in hyperspectral image classification task. To address the aforementioned problems, a deep convolutional network based multidimensional relation learning method is proposed in this paper. Firstly, we propose a weighted gaussian mask based spatial relation learning module to alleviate the spatial diffusion issue. Besides, we propose a channel attention alike spectral relation learning module to alleviate the interferences from noisy and redundant bands. Additionally, we propose an integrated decision based category relation learning module to alleviate the ill-conditioned classifying issue from imbalanced samples. Finally, we test the proposed method on four public and challenging datasets, and the experimental results validate the effectiveness and robustness of our method.

Our aim was to investigate factors predicting blood pressure (BP) variability during diagnostic cerebral angiography and associations between BP variability and clinical outcomes in patients with acute and subacute ischemic stroke and intracranial artery stenosis. 114 patients with ischemic stroke and intracranial artery stenosis (stenosis rate >50%) were recruited. Patients who underwent cerebral angiography within 3 days and 3–14 days of disease onset are referred to be Group A and Group S, respectively. BP variability in Group A was defined as the coefficient of variance (CV) of BP. Univariate and multivariate regression analyses were used to identify predictors of CV of BP and associations between CV of BP and clinical outcomes at discharge. In Group A patients, advanced age was associated with increased CV of SBP and diastolic blood pressure (DBP), and antihypertensive use was associated with lower CV of SBP. Male was associated with lower CV of DBP. In Group S, higher CV of SBP was associated with hypertension and antihypertensive use. Males had lower CV of SBP than females. The calcium channel blocker was associated with lower CV of DBP. Higher scores of the Stroke Scale at admission were significantly associated with poor clinical outcomes for both groups, while BP variability was not. Factors associated with BP variability are significantly different between stroke patients undergoing angiography within 3 days vs. 3–14 days after disease onset. BP variability is not significantly associated with clinical outcomes at discharge.

Objective Post-carotid endarterectomy hypertension is a well-recognized phenomenon closely related to surgical complications. This study aimed to determine whether different kinds of perioperative antihypertensive drugs had a protective effect on post-carotid endarterectomy hypertension and influence on intraoperative hemodynamics. Method We retrospectively investigated 102 carotid stenosis patients who underwent conventional endarterectomy with a perioperative baseline antihypertensive regimen. Post-carotid endarterectomy hypertension was defined as a postoperative peak systolic blood pressure ≥160 mmHg and/or a requirement for any additional antihypertensive therapies. We compared the clinical characteristics and types of baseline perioperative antihypertensive drugs between patients with and without post-carotid endarterectomy hypertension and then determined the significant independent effect of antihypertensive drugs on post-carotid endarterectomy hypertension through multivariate regression and detected their influence on intraoperative hypertension (induction-related systolic blood pressure and vasodilators consumption) and hemodynamic depression (intra-arterial systolic blood pressure ≤100 mmHg and/or heart rate ≤50 beats/min). We also investigated adverse events such as stroke, death, myocardial infarction, and cerebral hyperperfusion syndrome during the postoperative hospitalization. Results A total of 52/102 (51.0%) patients were defined as having post-carotid endarterectomy hypertension during the first three days postoperative, including eight patients with a postoperative systolic blood pressure that exceeded 160 mmHg at least once, 31 patients requiring postoperative antihypertensive treatment in addition to their baseline regimen, and 13 patients with both. The incidence of stroke/death/myocardial infarction and cerebral hyperperfusion syndrome after conventional endarterectomy during hospitalization were both 1.9%. A significantly increased risk of composite postoperative complications (including cerebral hyperperfusion syndrome, hyperperfusion-related symptoms, transient ischemic attacks, stroke, death, and cardiac complications) was observed in patients with post-carotid endarterectomy hypertension than without (15.4% versus 2.0%, p = 0.032). Patients free of post-carotid endarterectomy hypertension had a higher incidence of perioperative baseline β-blocker use than patients who suffered from post-carotid endarterectomy hypertension (46.0% versus 21%, p = 0.008). In multivariate analysis, β-blocker use was a significant independent protective factor for post-carotid endarterectomy hypertension (OR = 0.356, 95% CI: 0.146–0.886, p = 0.028). Patients taking β-blockers had a lower postoperative peak systolic blood pressure than the β-blocker-naïve population (137.1 ± 12.1 mmHg versus 145.0 ± 11.2 mmHg, p = 0.008), but the postoperative mean systolic blood pressure showed no intergroup difference. However, the incidence of hemodynamic depression during conventional endarterectomy was higher in patients with perioperative β-blocker use than in those without (44.1% versus 25.0%, p = 0.050). The difference in intraoperative hemodynamic depression became more prominent between the β-blocker and non-β-blocker groups (81.8% versus 33.3%, p = 0.014) for whose preoperative baseline heart rate was equal to or lower than 70 beats/min. Conclusion The perioperative use of β-blockers is a protective factor for post-carotid endarterectomy hypertension and contributes to stabilizing the postoperative peak systolic blood pressure three days after conventional endarterectomy. However, β-blockers might also lead to intraoperative hemodynamic depression, especially for patients with a low baseline heart rate.

Major depressive disorder (MDD) is a highly prevalent debilitating psychiatric disease and a serious public health problem worldwide. Brain structural MRI and postmortem studies on patients with depression have revealed changes in the anatomy and functionality in various brain regions, including the amygdala, thalamus, hippocampus, and prefrontal cortex (PFC). The alterations in these brain regions could be a result, in part, of the dysregulation of the neurotrophic factors. Neuregulin1 (NRG1) is one of the neurotrophic factors, and our previous study showed that the NRG1-ErbB4 signaling pathway plays a critical role in epilepsy. In this study, we established a chronic social defeat stress (CSDS) model to investigate the role of the NRG1-ErbB4 signaling pathway in depression-like behaviors. In CSDS mice, we found that the NRG1 protein expression levels were significantly decreased both in the medial prefrontal cortex (mPFC) and hippocampus, while phosphorylated ErbB4 only decreased in the mPFC. In addition, lateral ventricle NRG1 administration significantly rescued depression-like behaviors in the susceptible group. The current study suggests that the NRG1-ErbB4 signaling pathway may exert a protective role in MDD.

Lung cancer is the leading cause of cancer-related deaths worldwide. Owing to its high incidence and mortality, the development and discovery of novel anticancer drugs is of great importance. In recent years, many breakthroughs have been achieved in the search for effective anticancer substances from natural products. Many anticancer drugs used clinically and proven to be effective are derived from natural products. Quinonoids, including naphthoquinones, phenanthrenequinones, benzoquinones, and anthraquinones, constitute a large group of natural bioactive compounds that widely exist in higher and lower plant species. Given that most of these compounds possess anticancer effects, they are applied in many cancer studies, especially in lung cancer research. They can promote apoptosis, induce autophagy, and inhibit proliferation, angiogenesis, and cell invasion and migration. Some drugs can enhance anticancer effects when combined with other drugs. Thus, quinonoids have broad application prospects in the treatment of lung cancer. Here, we summarize the previous studies on the antilung cancer activities of quinonoids together with their underlying mechanisms and analyze the common research targets with different effects so as to provide references for the discovery of quinonoids against lung cancer.

BACKGROUND: Severe hyponatremia is considered a rare complication of pituitrin, which is widely used for the treatment of pulmonary hemorrhage. However, the management of pituitrin-associated hyponatremia can be challenging because a rapid correction of hyponatremia may cause the development of osmotic demyelination syndrome, resulting in life-threatening neurological injuries. CASE SUMMARY: A 20-year-old Chinese man with massive hemoptysis developed symptomatic hyponatremia (116 mmol/L) after therapy by a continuous intravenous drip of pituitrin. To normalize his serum sodium, a hypertonic saline infusion was applied for 3 d, and the pituitrin administration was stopped concurrently. Then, an overly rapid increase in serum sodium level (18 mmol/L in 24 h) was detected after treatment. One day later, the patient experienced a sudden onset of generalized tonic-clonic seizures, as well as subsequent dysarthria and dystonia. Magnetic resonance imaging revealed increased signal intensity in the bilateral symmetric basal ganglia on the T2-weighted images, compatible with a diagnosis of extrapontine myelinolysis. The patient received an intravenous administration of high-dose corticosteroids, rehabilitation, and neurotrophic therapy. Finally, his clinical abnormalities were vastly improved, and he was discharged with few residual symptoms. CONCLUSION: Physicians should be fully aware that pituitrin can cause profound hyponatremia and its correction must be performed at a controlled rate to prevent the development of osmotic demyelination syndrome.

Introduction Curcumin has various biological properties including being anti-inflammatory and antidiabetic. Podocyte apoptosis and autophagy dysfunction have been found to be responsible for the development of diabetic nephropathy (DN). Thus, the aim of the study was to investigate the effects of curcumin on the podocyte apoptosis and autophagy in DN and clarify its potential mechanisms. Methods The mice with DN induced by injection of streptozotocin were treated with curcumin by gavage at a dose of 200 mg/kg/day for 8 weeks. The serum lipid levels were detected by total cholesterol (TC) and triglyceride (TG) kits at different time points. Renal damage was assessed by detecting urine albumin, serum creatinine (Scr), HE staining and PAS staining. The renal impairment was detected by immunohistochemical staining and TUNEL staining. Western blot assay tested the expression of autophagy-related and apoptotic-related proteins in vivo and vitro. The viabilities and apoptosis of MPC5 cells exposed to high glucose (HG) or curcumin were respectively detected by CCK-8 assay and flow cytometry. Results The results showed that curcumin significantly decreased the progress of DN possibly via increasing autophagy and inhibiting apoptosis of renal cell in DN mice. Besides, podocyte marker proteins (podocalyxin and nephrin) were markedly increased in DN mice by curcumin treatment. The autophagy-related proteins LC3, p62, Beclin1, UVRAG and ATG5 were significantly affected in DN mice by curcumin, along with reducing expression of pro-apoptotic protein Bax and caspase-3 and increasing anti-apoptotic protein Bcl-2. In vitro, curcumin increased the viabilities and inhibited apoptosis of MPC5 cells exposed to high glucose (HG). In addition, the podocyte autophagy was enhanced partly via regulating beclin1/UVRAG. Discussion Together, the results showed that curcumin inhibited podocyte apoptosis and accelerated cell autophagy via regulating Beclin1/UVRAG/Bcl2. Thus, the study showed that curcumin exerted significantly protective effects in DN.

Background The current work aimed to assess whether Gynostemma pentaphyllum (GP), a Chinese herbal medicine, structurally modifies the gut microbiota in rats during non-alcoholic fatty liver disease (NAFLD) treatment. Methods High-fat diet (HFD)-induced NAFLD rats were orally administered water decoction of GP or equal amounts of distilled water per day for 4 weeks. Liver tissues were examined by histopathological observation, while intestinal tissues were examined by both histopathological and ultrastructural observations. The levels of fasting blood glucose (FBG), fasting serum insulin (FINS), total cholesterol (TC), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), alanine transaminase (ALT) and aspartate transaminase (AST) were measured by enzymatic method. The levels of toll-like receptor 4 (TLR-4), tumor necrosis factor-alpha (TNF-α), interleukin-1-beta (IL-1β) and interleukin-6 (IL-6) in both serum and hepatic tissues were measured by RT-qPCR. The protein expression level of TLR-4 in hepatic tissues was detected by western blot. The gut microbiota was assessed by 16S rRNA-based microbiota analysis. Results GP maintained intestinal integrity and reversed gut dysbiosis in high-fat diet (HFD)-induced NAFLD rats. This also reduced the ratio of Firmicutes to Bacteroidetes, enriching the abundance of beneficial bacteria (Lactococcus spp.) and inhibiting the abundance of pathogenic bacteria (Ruminococcus spp.) in the gut. The levels of pro-inflammatory cytokines (TNF-α, IL-1β and IL-6) and the expression of TLR4 were downregulated (P ), while the insulin resistance index, HOMA-IR showed improvement by GP treatment (P ). Liver function indicators (ALT and AST) were remarkably decreased (P ). Besides, GP treatment reduced TG and LDL-C levels (P ), and increased HDL-C level (P ) compared with NAFLD group. Conclusion The structural alterations of gut microbiota induced by GP are associated with NAFLD alleviation.

Finding treatments that accelerate peripheral nerve regeneration, prolongation, and functional recovery remains a challenging task. Platelet-rich plasma (PRP) contains numerous growth factors and active proteins, and low-dose ultrashort waves (USWs) stimulate the formation of nerve-nourishing vessels, which are powerful for nerve regeneration. The goal of this study was to evaluate the synergistic effects of serial ultrasound-guided PRP injections combined with low-dose USWs radiation on peripheral nerve regeneration in a crush injury model. Fifty rabbits were equally and randomly divided into normal control, model, USW, PRP, and PRP+USW groups. The neurological function, electrophysiological recovery, and histological and morphological evaluation of regenerated nerves, as well as a targeted muscle recovery assessment, were performed to investigate the regenerative effect of PRP combined with USW therapy. Our results showed that the PRP+USW group had the better early axonal regeneration and displayed the earliest positive compound muscle action potential among the treatment groups. At postintervention week 12, a histological evaluation showed similar expression of the S-100 protein in the PRP+USW and normal control groups. Moreover, the morphological assessment revealed a significant increase in the myelinated nerve fiber density and diameter and myelin sheath thickness compared with the USW and PRP groups. The morphometry of the target muscles indicated the lowest reduction in the percent volume in the PRP+USW group, and an ultrasound examination of the targeted muscle showed the best improvement in stiffness and perfusion parameters at 12 weeks after crush injury. Thus, serial ultrasound-guided PRP injections combined with low-dose USW radiation exert a synergistic effect on accelerating functional axon recovery and decreasing atrophy of the target muscles in a crush injury model. Impact Statement This research describes that the application of platelet-rich plasma combined with low-dose ultrashort waves treatment exert a synergistic effect on accelerating peripheral nerve regeneration. With the extensive use of platelet-rich plasma and physical factors in regenerative medicine or clinical rehabilitation medicine, our findings may help establish effective strategies for repairing peripheral nerve injury.

Mitochondrial genes' variants encoded in both the nuclear and mitochondrial genomes can disrupt mitochondrial function, resulting in losing of cartilage and generating osteoarthritis (OA). However, the association between mtDNA haplogroups and OA still lacks strength evidence supporting. The aim of this meta-analysis is to assess the role of mtDNA haplogroups in speculating the pathogenesis and progression of OA. PubMed, Embase, the Cochrane Central Register of Controlled Trials, and World Health Organization clinical trials' registry center were searched to identify relevant studies up to the end of March 2019. Inclusion citations required a case-control or cohort study to demonstrate the association between mtDNA haplogroups and OA's prevalence or progression. Title, abstract, and full-text screening were sequentially assessed by three reviewers. Data were analyzed using STATA. Besides, publication bias and meta-regression analysis were conducted to explore potential heterogeneities. We collected results from 7 articles. The cluster TJ cases showed a lower proportion in OA cases (RR = 0.83, 95% CI 0.72, 0.96). However, there is no evidence that revealed this kind of impact originated from neither type J nor type T individually. Besides, the type B and G analyses among Asian populations also elucidated a negative association. Moreover, the cluster TJ of mtDNA haplogroups revealed a lower cumulative probability of radiographic OA progression (ES = 0.77, 95% CI 0.63, 0.94), which was contributed by type T (ES = 0.61, 95% CI 0.45, 0.82).The mtDNA haplogroups do have impacts on the prevalence and progression of OA. Cluster TJ could help reduce the prevalence and slow down the radiographic changes; however, the impacts came from type J and type T, respectively.

Purpose The purpose of this study is to retrospect and summarize clinical efficacy and experience of the free perforator flap base on the superficial palmar branch of the radial artery for tissue defect reconstruction in hand. Method 17 patients who underwent tissue defect in hands reconstruction by the free superficial palmar branch of the radial artery (SPBRA) perforator flaps in our department from July 2014 to October 2018 were reviewed. Results All the flaps in our series application were survival uneventful except one, which was necrosis because of venous thrombosis postoperative 5 days, and then the abdominal pedicle flap was executed to recover the defect in second stage. The first dorsal metacarpal artery flap and the arterial venous flap were utilized to cover the defect in one right index finger and one right ring finger due to the absence variation of the SPBRA. 2 cases presented tension vesicle of superficial skin and 1 case occurred venous congestion. All donor sites were closed primarily. The follow-up period means 13.5 months (range, 4–50 months). The static 2 point discrimination test mean 7.53 mm (range, 4–11 mm). All flaps acquire protective feeling at the latest follow-up. The self-assessment of patients: 13 cases in good, 4 cases in fair. Conclusion The goal of physiological reconstruction and esthetic effect can be achieved for hand tissue defect by the free SPBRA perforator flap, multiple tissues of the flap can be contained according to the defect. Even though the SPBRA is variation, arterial venous flap could be applied thanks to abundant superficial cutaneous veins.

Accurate traffic forecasting on citywide networks is one of the crucial urban data mining applications that accurately provide congestion warning and transportation scheduling. While previo...

Deep learning on point clouds has attracted increasing attention in the fields of 3D computer vision and robotics. In particular, graph-based point-cloud deep neural networks (DNNs) have demonstrated promising performance in 3D object classification and scene segmentation tasks. However, the scattered and irregular graph-structured data in a graph-based point-cloud DNN cannot be computed efficiently by existing SIMD architectures and accelerators. We present Point-X, an energy-efficient accelerator architecture that extracts and exploits the spatial locality in point cloud data for efficient processing. Point-X uses a clustering method to extract fine-grained and coarse-grained spatial locality from the input point cloud. The clustering maps the point cloud into distributed compute tiles to maximize intra-tile computational parallelism and minimize inter-tile data movement. Point-X employs a chain network-on-chip (NoC) to further reduce the NoC traffic and achieve up to 3.2 × speedup over a traditional mesh NoC. Point-X’s multi-mode dataflow can support all common operations in a graph-based point-cloud DNN, i.e., edge convolution, shared multi-layer perceptron, and fully-connected layers. Point-X is synthesized in a 28nm technology and it demonstrates a throughput of 1307.1 inference/s and an energy efficiency of 604.5 inference/J on the DGCNN workload. Compared to the Nvidia GTX-1080Ti GPU, Point-X shows 4.5 × and 342.9 × improvement in throughput and efficiency, respectively.

Deep learning on point clouds has attracted increasing attention in the fields of 3D computer vision and robotics. In particular, graph-based point-cloud deep neural networks (DNNs) have demonstrated promising performance in 3D object classification and scene segmentation tasks. However, the scattered and irregular graph-structured data in a graph-based point-cloud DNN cannot be computed efficiently by existing SIMD architectures and accelerators. Following a review of the challenges of point-cloud DNN and the key edge convolution operation, we provide several directions in optimizing the processing architecture, including computation model, data reuse, and data locality, for achieving an effective acceleration and an improved energy efficiency.

Evidence supports linezolid therapeutic drug monitoring as the exposure–response relationship has been identified for toxicity among patients receiving linezolid, but the data to establish the upper limit are limited and the published toxicity thresholds range widely. The purpose of this study was to determine the linezolid exposure–toxicity thresholds to improve the safety of linezolid. This is a multicenter retrospective study of adult patients treated with linezolid from 2018 to 2019. The population pharmacokinetic model of linezolid was established based on 270 plasma concentrations in 152 patients, which showed creatinine clearance and white cell count are covariates affecting the clearance of linezolid, and serum albumin is the covariate affecting the volume of distribution. Classification and regression tree analysis was used to determine the linezolid exposure thresholds associated with an increased probability of toxicity. Among 141 patients included for toxicity analysis, the rate of occurring toxicity was significantly higher among patients with an AUC0-24, d1 ≥163 mg h/L, AUC0-24, d2 ≥207 mg h/L, AUC0-24, ss ≥210 mg h/L, and Cmin,d2 ≥6.9 mg/L, Cmin,ss ≥6.9 mg/L, while no threshold was discovered for Cmin, d1. Those exposure thresholds and duration of linezolid treatment were independently associated with linezolid-related toxicity in the logistic regression analyses. In addition, the predictive performance of the AUC0-24 and Cmin thresholds at day 2 and steady state were close. Considering that the AUC estimation is cumbersome, Cmin threshold at 48 h and steady state with a value of ≥6.9 mg/L is recommended to improve safety, especially for patients with renal insufficiency and patients with low serum albumin.

Purpose: A low ABI, ≦0.9, indicates peripheral artery disease (PAD) and physical activity (PA) represents an important non-surgical treatment for patients with PAD. However, as for the general population, the associations between PA, PAD, and their mutual dependence are not well-defined. Here we aimed to determine whether there is a dose–response relationship between PA and incidence of PAD in the general population using restricted cubic spline (RCS).Patients and methods: This study analyzed 1,370 adults aged ≧40 years who had participated in the National Health and Nutrition Examination Survey (NHANES) during 1999–2004. The ABI of the participants were measured by trained technicians, and PAD was defined as ABI ≦0.9. PA was obtained with a standard questionnaire, and metabolic equivalents (MET) were used to quantify the PA level. Logistic regression was used to assess the association between PA and incidence of PAD, and the dose–response relationship was analyzed with RCS.Results: PAD was present in 6.2% of the participants: 5.6% of males and 6.9% of females. After adjusting for potential confounders, compared with the first quartile (Q1) of MET, the odds ratios (ORs) of PAD for those with Q2, Q3, and Q4 of MET were 0.688 [95% confidence interval (CI) = 0.684–0.692], 0.463 (95% CI = 0.460–0.466), 0.816 (95% CI = 0.812–0.821), respectively (all p < 0.0001). The RCS regression showed that physical activity was related to the incidence of PAD in a non-linear manner (p for non-linearity < 0.0001). For females, the prevalence of PAD decreased as physical activity increased, reaching the minimum for activity at ~5,800 MET-min month−1 (OR = 0.425, 95% CI = 0.424–0.426), and for males, no plateau was found in this study.Conclusion: The prevalence of PAD is inversely associated with PA, and vigorous activities might help decrease PAD risk for general population. The prevalence of PAD reaches the minimum at ~5,800 MET-min month−1, representing a recommended PA value.

Three independent Ti-oxo clusters (TOCs) that contain 6, 8, and 12 Ti atoms in the cores and alkyl groups on the surface were developed as cathode interlayers in bulk-heterojunction organic solar cells (OSCs). These TOCs have precise chemical structures with a single crystal, excellent solubility in methanol, and well-aligned work function. Smooth films can be facilely obtained by spin-casting their solution on top of the active layer. Therefore, they can be used as an interlayer in OSCs to provide a high power conversion efficiency (17.29%). Further studies reveal that these TOCs can not only reduce the work function of the silver electrode to provide better energy level alignment but also exhibit a significant n-doping effect with the non-fullerene acceptors to facilitate efficient electron extraction and transport. Our results demonstrate that TOCs as semiconductors have great potential application in OSCs.

Studies have shown that addictive behavior is associated with many brain regions, such as the insula, globus pallidus, amygdala, nucleus accumbens, and midbrain dopamine system, but only a few studies have explored the role of the dorsal striatum in addictive behavior. In June 2020, we started contacting 608 patients who were hospitalized between January 2017 and December 2019, and we recruited 11 smoking addicts with dorsal striatum damage and 20 controls with brain damage that did not involve the dorsal striatum (the damaged areas included the frontal lobe, temporal lobe, parietal lobe, brain stem, thalamus, internal capsule, and so on). All study participants had brain damage due to acute cerebral infarction. Disruption of smoking addiction was found to be significantly associated with the dorsal striatum (Phi = 0.794770, P = 0.000015). Our findings suggested that patients in the dorsal striatum group were more likely to discontinue smoking than those in the non-dorsal striatum group. The characteristics of this interruption is that smoking can be quit more easily and quickly without recurrence and that the impulse to smoke is reduced. These results suggest that the dorsal striatum is a key area for addiction to smoking.

We report a case of community-acquired pneumonia in a patient in China. We verified Legionella pneumophila infection through next-generation sequencing of blood, sputum, and pleural effusion samples. Our results show the usefulness of next-generation sequencing and of testing different samples early in the course of illness to identify this bacterium.

Recently, progress on semantic image segmentation is substantial, benefiting from the rapid development of Convolutional Neural Networks. Semantic image segmentation approaches proposed lately have been mostly based on Fully convolutional Networks (FCNs). However, these FCN-based methods use large receptive fields and too many pooling layers to depict the discriminative semantic information of the images. Specifically, on one hand, convolutional kernel with large receptive field smooth the detailed edges, since too much contexture information is used to depict the “center pixel.” However, the pooling layer increases the receptive field through zooming out the latest feature maps, which loses many detailed information of the image, especially in the deeper layers of the network. These operations often cause low spatial resolution inside deep layers, which leads to spatially fragmented prediction. To address this problem, we exploit the inherent multi-scale and pyramidal hierarchy of deep convolutional networks to extract the feature maps with different resolutions and take full advantages of these feature maps via a gradually stacked fusing way. Specifically, for two adjacent convolutional layers, we upsample the features from deeper layer with stride of 2 and then stack them on the features from shallower layer. Then, a convolutional layer with kernels of 1× 1 is followed to fuse these stacked features. The fused feature preserves the spatial structure information of the image; meanwhile, it owns strong discriminative capability for pixel classification. Additionally, to further preserve the spatial structure information and regional connectivity of the predicted category label map, we propose a novel loss term for the network. In detail, two graph model-based spatial affinity matrixes are proposed, which are used to depict the pixel-level relationships in the input image and predicted category label map respectively, and then their cosine distance is backward propagated to the network. The proposed architecture, called spatial structure preserving feature pyramid network, significantly improves the spatial resolution of the predicted category label map for semantic image segmentation. The proposed method achieves state-of-the-art results on three public and challenging datasets for semantic image segmentation.

This paper presents a comparative analysis between two new architectures for RF programmable-gain amplifiers (RFPGAs): voltage-mode RFPGA-V and current-mode RFPGA-I. RFPGA-V utilizes multiple-switch-multiple-amplifier configuration and gain interpolation method to achieve a fine gain step of 0.25-dB over 42-dB gain range for the band of 250 MHz to 2.3 GHz. Meanwhile, RFPGA-I uses a current steering approach to achieve a fine gain step of 0.25-dB over 42-dB gain range for an even wider band of 250 MHz to 3.4 GHz. Since the active feedback topology is used, no off-chip inductor is needed in either RFPGA, especially for the low-frequency band. In addition, both RFPGA-V and RFPGA-I are able to handle maximum 4.4 V peak-to-peak input signal without compromising their high operating bandwidth. Two broadband RF front ends for direct sampling receivers, which include either RFPGA-V or RFPGA-I followed by the same gain buffer and RF filter, have been demonstrated. For the RF front end with RFPGA-V, the measured gain, noise figure, third-order input-referred intercept point (IIP3), and second-order input-referred intercept point (IIP2) in the differential mode are 29.5 dB, 5.2 dB, −10.9 dBm, and 31.4 dBm, respectively. With RFPGA-I, they are 30.5 dB, 3 dB, −10.5 dBm, and 21.1 dBm, respectively. Both RF front ends consume approximated 50 mW and occupy the similar area of 0.32 mm2 in 28-nm CMOS technology.

Summary Background & aims Vitamin B6 has been postulated to play an important role in determining chronic diseases. However, few studies have evaluated associations between dietary vitamin B6 and cause-specific mortality comprehensively. Methods We investigated the associations between vitamin B6 from diet and risk of all-cause, and cause-specific mortality in 134,480 participants from the Shanghai Men's Health Study (2002–2014) and Shanghai Women's Health Study (1997–2014). The median follow-up periods for men and women were 10.3 and 16.2 years, respectively. We estimated hazard ratio (HR) and 95% confidence interval (CI) using Cox proportional hazards models. Results After adjustment for suspected confounders, the multivariable-adjusted HRs for the highest versus lowest quintiles for total, CVD, stroke and CHD mortality among men were 0.83 (95%CI = 0.76, 0.90), 0.73 (95%CI = 0.63, 0.85), 0.71 (95%CI = 0.58, 0.88), 0.66 (95%CI = 0.47, 0.91), accordingly. Women with the highest intake had significantly 17% (HR = 0.83; 95% CI = 0.77, 0.90), 20% (HR = 0.80; 95% CI = 0.70, 0.92), and 28% (HR = 0.72; 95% CI = 0.59, 0.86) lower risks of total, CVD and stroke mortality compared with those of women with lowest vitamin B6 intake. No significant association was observed between dietary vitamin B6 and cancer mortality both among men and women. Conclusions In the current study with two prospective Chinese cohorts, high dietary vitamin B6 consumption was inversely associated with risk of all-cause and CVD mortality.

The aim of the present study was to identify predictive factors for cervical cancer (CC) progression using a multistage approach. The present study obtained data from 390 healthy women and 259 patients with cervical cancer between June 2012 and June 2017, and used a multiple stage re-analysis strategy for clinical detection of CC. A total of seven types of serum indices were used in the present study, including sugar chain antigen 125 (CA-125), sugar chain antigen 199 (CA-199), α fetoprotein (AFP), carcino- embryonic antigen, alkaline phosphatase (ALP), cholesterol and triglyceride (TG). The expression levels of CA-125, CA-199, AFP, ALP, cholesterol and TG were significantly different between healthy women and patients with cervical squamous cell carcinoma (SCC). Furthermore, ALP, cholesterol and TG expression levels were significantly different in healthy women compared with patients with cervical adenocarcinoma (AC). Further comparisons based on age and pathological staging demonstrated that the variability in the ALP level was not significant between the 0.05); however, was significant in patients with SCC (P

Background This study prospectively investigated the efficacy and radiation dose of ultralow dose computed tomography (CT)‐guided hook‐wire localization (HWL) at 100 kV with tin filtration (100Sn kV) for small solitary pulmonary nodules. Methods All HWL procedures were performed on a third generation dual‐source CT system. Eighty‐eight consecutive patients undergoing CT‐guided HWL were randomly assigned to standard dose CT (Group A: n = 44; reference 110 kV and 50 mA) or ultralow dose CT (Group B: n =44; 100Sn kV and 96 mA) protocols. The technical success rate, complications, subjective image quality, and radiation dose were compared between the groups. Results The mean volume CT dose index and total dose‐length product were significantly lower in Group B compared to Group A (0.32 mGy vs. 3.2 ± 1.1 mGy and 12.1 ± 0.97 mGy‐cm vs. 120 ± 40.6 mGy‐cm; P < 0.001). The effective dose in Group B was significantly lower than in Group A (0.17 ± 0.01 mSv vs. 1.68 ± 0.57 mSv, ‐89.8%; P < 0.001). The technical success rates were 100% for both groups. There were no significant differences in complication rates between the protocols (P > 0.05). The image quality of ultralow dose CT met the requirements for HWL procedure. Conclusion Ultralow dose CT‐guided HWL of solitary pulmonary nodules performed at 100Sn kVp spectral shaping significantly reduced the radiation dose compared to standard dose CT, with high technical success and acceptable patient safety.

Purpose: To investigate antimicrobial resistance and risk factors for mortality of Klebsiella pneumoniae (KP) pneumonia in diabetics and nondiabetics. Patients and methods: A retrospective study was conducted among inpatients of KP pneumonia via electronic medical records in a territory hospital between January 2016 and June 2018. Antimicrobial resistance in KP pneumonia was compared between diabetics and nondiabetics. Independent risk factors for mortality in KP pneumonia were identified by univariate and multivariate logistic regression among diabetics and nondiabetics separately. Results: In this study, 456 patients with KP pneumonia were included. There were 156 cases with diabetes and 300 without diabetes. KP showed a lower antimicrobial resistance to a multitude of antimicrobials in pneumonia among diabetics than nondiabetics, namely aztreonam, cefotetan, sulperazone, meropenem, amikacin, tobramycin, sulfamethoxazole, and fosfomycin. In addition, carbapenem-resistant Klebsiella pneumoniae (CRKP) was more prevalent among nondiabetics than diabetics who were admitted to intensive care unit (ICU) (63.0% vs 45.1%, P = 0.038). Multivariable analysis showed that independent risk factors for in-hospital mortality (IHM) in KP pneumonia among diabetics differed from that among nondiabetics as well. Independent predictors for IHM of KP pneumonia among diabetics were male (OR: 5.89, 95% CI: 1.34-25.93, P = 0.019), albumin (ALB) < 35 g/L (OR: 7.00, 95% CI: 2.02-24.28, P = 0.002), bloodstream infection (BSI) (OR: 21.14, 95% CI: 3.18-140.72, P = 0.002), and invasive ventilation during hospitalization (OR: 8.00, 95% CI: 2.99-21.42, P < 0.001). In nondiabetics, independent predictors were higher CURB-65 score (OR: 1.92, 95% CI: 1.29-2.86, P = 0.001), CRKP (OR: 2.72, 95% CI: 1.07-6.90, P = 0.035), BSI (OR: 4.98, 95% CI: 1.34-18.50, P = 0.017), and ICU admission (OR: 4.06, 95% CI: 1.57-10.47, P = 0.004). Conclusion: In KP pneumonia, diabetics showed lower antimicrobial resistance and different independent risk factors for mortality compared with nondiabetics, in line with previous studies. Importantly, further attention should be paid on rational and effective antibiotic and supportive treatments in order to reduce mortality without aggravating antimicrobial resistance and metabolic damage among diabetics.

Background. The Traditional Chinese Medicine (TCM) Tian Wang Bu Xin Dan (TWBXD) has been used widely for treating insomnia in China. The purpose of this meta-analysis was to evaluate the efficacy and safety of TWBXD in the treatment of insomnia. Objective. This study evaluated the efficacy and safety of TWBXD for insomnia. Methods. We searched seven main databases including PubMed, EMBASE, Cochrane Central Register of Controlled Trials, Chinese Biomedicine Database (CBM), China National Knowledge Infrastructure (CNKI), Chinese Scientific Journals Database (VIP), and Wan-fang. We identified randomized-controlled trials (RCTs) for insomnia treatment involving TWBXD, TWBXD combined with conventional Western medicine, and conventional Western medicine from their inception to May 2018. The quality of literature was evaluated by Cochrane assessing tool to reduce the risk of bias. Meta-analysis and heterogeneity of results across the trials were performed. RevMan 5.3 was used to synthesize the results. Results. 14 studies involving 1,256 participants were identified in this systematic review. Methodological deficiencies existed in most of the included trials. Few studies described the generation of a random sequence in detail, the concealment of allocation, and the methods of blinding. No placebo was used in treatment. 12 trials compared TWBXD with conventional Western medicine and 2 trials compared TWBXD combined with conventional Western medicine. The results of our meta-analysis showed relative benefits in effective rates in favor of TWBXD (Odds Ratio [OR] 2.71, 95% confidence interval [CI] 1.67 to 4.39, P < 0.00001) and TWBXD combined with conventional Western medicine (OR 5.05, 95% CI 1.58 to 16.12, P=0.006). The Pittsburgh Sleep Quality Index (PSQI) scores showed similar results, which favored TWBXD (Weighted Mean Difference [WMD] -1.82, 95% CI -3.00 to -0.64, P=0.003). Only 5 trials reported adverse events, whereas the other 9 trials did not provide the safety information. Conclusion. This review demonstrates that although the effects of TWBXD on insomnia were promising, they need to be interpreted with caution, due to the poor methodological quality and the small number of trials of the included studies. TWBXD seems to be generally safe, but there is insufficient evidence to make conclusions on the safety because fewer studies reported the adverse events. Further studies on a larger scale with more rigorous designs are required to evaluate the role of TWBXD in the insomnia treatment.

Automatic gender and age prediction has become relevant to an increasing amount of applications, particularly under the rise of social platforms and social media. However, the performances of existing methods on real-world images are still not satisfactory as we expected, especially when compared to that of face recognition. The reason is that, facial images for gender and age prediction have inherent small inter-class and big intra-class differences, i.e., two images with different skin colors and same age category label have big intra-class difference. However, most existing methods have not constructed discriminative representations for digging out these inherent characteristics very well. In this paper, a method based on muti-stage learning is proposed: The first stage is marking the object regions with an encoder-decoder based segmentation network. Specifically, the segmentation network can classify each pixel into two classes, “people” and others, and only the “people” regions are used for the subsequent processing. The second stage is precisely predicting the gender and age information with the proposed prediction network, which encodes global information, local region information and the interactions among different local regions into the final representation, and then finalizes the prediction. Additionally, we evaluate our method on three public and challenging datasets, and the experimental results verify the effectiveness of our proposed method.

Remote sensing image scene classification is a fundamental problem, which aims to label an image with a specific semantic category automatically. Recently, deep learning methods have achieved competitive performance for remote sensing image scene classification, especially the methods based on a convolutional neural network (CNN). However, most of the existing CNN methods only use feature vectors of the last fully connected layer. They give more importance to global information and ignore local information of images. It is common that some images belong to different categories, although they own similar global features. The reason is that the category of an image may be highly related to local features, other than the global feature. To address this problem, a method based on rearranged local features is proposed in this paper. First, outputs of the last convolutional layer and the last fully connected layer are employed to depict the local and global information, respectively. After that, the remote sensing images are clustered to several collections using their global features. For each collection, local features of an image are rearranged according to their similarities with local features of the cluster center. In addition, a fusion strategy is proposed to combine global and local features for enhancing the image representation. The proposed method surpasses the state of the arts on four public and challenging data sets: UC-Merced, WHU-RS19, Sydney, and AID.