Search

With the powerful learning ability of deep convolutional networks, deep clustering methods can extract the most discriminative information from individual data and produce more satisfactory clustering results. However, existing deep clustering methods usually ignore the relationship between the data. Fortunately, the graph convolutional network can handle such relationship, opening up a new research direction for deep clustering. In this paper, we propose a cross-attention based deep clustering framework, named Cross-Attention Fusion based Enhanced Graph Convolutional Network (CaEGCN), which contains four main modules: the cross-attention fusion module which innovatively concatenates the Content Auto-encoder module (CAE) relating to the individual data and Graph Convolutional Auto-encoder module (GAE) relating to the relationship between the data in a layer-by-layer manner, and the self-supervised model that highlights the discriminative information for clustering tasks. While the cross-attention fusion module fuses two kinds of heterogeneous representation, the CAE module supplements the content information for the GAE module, which avoids the over-smoothing problem of GCN. In the GAE module, two novel loss functions are proposed that reconstruct the content and relationship between the data, respectively. Finally, the self-supervised module constrains the distributions of the middle layer representations of CAE and GAE to be consistent. Experimental results on different types of datasets prove the superiority and robustness of the proposed CaEGCN.

Many real-world applications can be formulated as expensive multimodal optimization problems (EMMOPs). When surrogate-assisted evolutionary algorithms (SAEAs) are employed to tackle these problems, they not only face the problem of selecting surrogate models but also need to tackle the problem of discovering and updating multiple modalities. Different optimization problems and different stages of evolutionary algorithms (EAs) generally require different types of surrogate models. To address this issue, in this article, we present a multisurrogate-assisted multitasking particle swarm optimization algorithm to seek multiple optimal solutions of EMMOPs at a low computational cost. The proposed algorithm first transforms an EMMOP into a multitasking optimization problem by integrating various surrogate models, and designs a multitasking niche particle swarm algorithm to solve it. Following that, a surrogate model management strategy based on the skill factor and clustering is developed to effectively balance the number of real function evaluations and the prediction accuracy of candidate optimal solutions. In addition, an adaptive local search strategy based on the trust region is proposed to enhance the capability of swarm in exploiting potential optimal modalities. We compare the proposed algorithm with five state-of-the-art SAEAs and seven multimodal EAs on 19 benchmark functions and the building energy conservation problem and experimental results show that the proposed algorithm can obtain multiple highly competitive optimal solutions.

Human Activity Recognition(HAR) based on Channel State Information(CSI) plays an increasingly important role in the research of human-computer interaction. Many CSI HAR models based on traditional machine learning methods and deep learning methods have encountered two challenges. A lot of CSI activity data is needed to train the HAR models, which is time-consuming. When the indoor environment or scene changes, the recognition accuracy of the model drops significantly, so it is necessary to re-collect data to train the model. The existing few-shot learning-based method can solve the above problems to some extent, but when there are more kinds of new activities or fewer shots, the recognition accuracy will decrease significantly. In this paper, considering the relationship between various activity data, a Graph-based few-shot learning mothod with Dual Attention Mechanism(CSI-GDAM) is proposed to perform CSI-based HAR. The model uses a feature extraction layer including Convolutional Block Attention Module(CBAM) to extract activity-related information in CSI data. The difference and inner product of the feature vector of the CSI activity samples are used to realize the graph convolutional network with graph attention mechanism. The experiments proved that under the learning task of recognizing new activities in the new environment, the recognition accuracy rates reached 99.74% and 98.42% in the 5-way 5-shot and 5-way 1-shot cases, respectively. The proposed method is also compared with other few-shot learning and transfer learning methods.

This paper presents a physical model for monolithic self-rolled-up microtube interdigital capacitors and elaborates their working mechanisms in comparison to on-chip planar interdigital capacitors. Besides the high-frequency phenomena such as the skin effect in thin, wide metal films, the model accounts for the complexities resulting from all possible 3D electrode configurations, such as the inner diameter dependent overlapping capacitances and edge capacitances between electrodes, as well as parasitic inductances within the rolled-up structure between the fingers, for integer and non-integer number of turns. The model is validated by the measured results of fabricated self-rolled-up membrane (S-RuM) interdigital capacitors over a wide range of layout and process parameters. This fully enables the structural design optimization for electrical performance of this 3D device architecture.

Background: Atrial fibrillation (AF) is the most common persistent arrhythmia and an important factor leading to cardiovascular morbidity and mortality. Several key genes and diagnostic markers have been discovered with the development of advanced modern molecular biology techniques, but the etiology and pathogenesis of AF remained unknown. Methods: In this study, three chip-seq data sets and a RNA-seq data set were integrated as a comprehensive network for pathway analysis of the biological functions of related genes in AF, hoping to provide a better understanding on the etiology and pathogenesis of AF. Results: Differential co-expression analysis identified 360 genes with specific expression in AF, and functional enrichment analysis further revealed that these genes were significantly correlated with focal expression (p 0.9) and in different platform data sets (mean AUC>0.75). Conclusion: Finally, a diagnostic model for AF established, thus showing its potential in the early identification and prediction of AF.

Graphical abstract, Highlights • The mappings between micro cracks and diffraction pattern curves were revealed based on GIXD and MD. • Surface and subsurface defects of CaF2 crystals induced by abrasive machining were evaluated and characterized accurately. • MD simulated results agreed well with the experimental results. • This work provided a novel technology for nondestructive testing of defects of single crystals at nano- and micro- scales., Introduction Surface and subsurface defects were easily induced during abrasive machining process of optical crystals due to their high brittleness. Accurate characterization of these defects is the prerequisite for obtaining optical components with high surface integrity. Objectives This work aims to evaluate subsurface defects of CaF2 single crystals induced by abrasive machining, the mappings between micro cracks and diffraction pattern curves, and the influence of micro cracks on lattice structures. Methods Molecular dynamics simulation, grazing incidence X-ray diffraction experiments and cross-sectional TEM detection were used in this work. Results In grazing incidence X-ray diffraction (GIXD) detection experiments on lapping and polished specimens, shifts phenomenon of the peak position under the non-theoretical grazing incidence angle indicated that the subsurface was damaged to a certain extent. The micro cracks of the subsurface were evaluated by the consistent characteristic of “peak drift” of the diffraction pattern curve in both experiments and simulations. In addition, cross-sectional TEM results showed that regular micro cracks were found on the subsurface, which agreed well with the simulation results. Conclusion The subsurface defects of CaF2 single crystals induced by abrasive machining, and the influence of micro cracks on lattice structures can be evaluated by molecular dynamics simulation. The simulation results revealed the mappings between the micro cracks and diffraction pattern curve, which demonstrated that a phenomenon of “peak drift” occurred near the diffraction angle of a specific crystal plane. This work provided a novel technology for the nondestructive testing of defects of single crystal materials at nano- and micro- scales.

‘Co-control’, or synergistically reducing CO2 and local air pollutant emissions, is an important strategy for cities to achieve ‘low carbon’ and ‘blue sky’ simultaneously. However, there were few studies to evaluate and compare the level of co-control of CO2 and local air pollutants in cities yet. The present study proposed qualitative and quantitative methods to evaluate the level of co-control of CO2 and three local air pollutant (SO2, NOx, and particulate matter PM) emissions in key environmental protection cities in China over two periods (2012–2015 and 2015–2018). Statistical analysis found that, though three local air pollutant emissions positively correlated with CO2 emission, no significantly positive correlation was found between local air pollutants emission reductions and CO2 emission reductions, indicating that co-control effects were poor in general. By using the co-control effect coordinate system, qualitative evaluation showed that less than half of the sample cities could achieve co-control of the total amount of CO2 and local air pollutants. By employing the indicator of emission reduction equivalence (EReq), quantitative evaluation showed that the co-control level of the sample cities improved in 2015–2018 compared with 2012–2015. Further regression analysis showed that, reducing coal consumption and economic development significantly enhanced the co-control performance of the observed cities. The present case study proved that the proposed methods for evaluation and comparison of the city co-control performance works well, and can be used in other countries and regions to promote global cities racing to carbon and local air pollutants co-control.

A novel waveguide-to-microstrip inline transition was proposed using a wedge-waveguide iris to form an ultra-compact U-bend waveguide. The wedge-waveguide iris not only functioned as a wide wall of the height-reduced waveguide for the E-plane probe, but also played a critical role in impedance matching and parasitic resonance suppression. The working principle was investigated through theoretical analysis and full-wave simulation, and the relevant design processes for the transition were discussed. Moreover, two back-to-back transition prototypes, one working at the W-band and the other working at the WR4.3 band, were designed, fabricated, and measured. The back-to-back prototypes obtained an insertion loss of less than 0.8 and 1.2 dB, with a return loss better than 15 dB in the W-band and WR4.3 band, respectively. The measurement results were consistent with simulation results. A miniaturized 110-GHz tripler incorporating this transition for inline output was fabricated. A conversion efficiency of 2.5%-3.6% was achieved in 90-120 GHz at 100-mW driving power.

On-line detection of fault characteristic gases released by the insulation oil of transformers is developed as a scientific basis for estimating explosion accidents. However, the performance of most gas sensors, such as H2 and C2H2 sensors, is usually affected by temperature, and there exists large detection error when the temperature of transformer oil changes in operation. And the existing temperature compensation methods suffer from problems such as low integration, large volume, and high costs. Herein, a sensor array based on three triple-electrode sensors with carbon nanotubes was fabricated. Each of the three sensors has three vertically stacked electrodes with various interelectrode distances. The first and second sensors were used to detect H2 and C2H2, respectively, and the third sensor was used to measure temperature. All the detection mechanisms are based on the dependence of discharge current on gas concentration and temperature. The sensor array could detect H2/C2H2 mixture and temperature due to the exponential function between the output currents and electrode separations. The proposed sensor exhibited low relative and repeatability error values less than 5.1% and 4.9%, respectively, and displayed humidity compensation potential like temperature. High integration makes the sensor array superior among the other mechanism sensors.

In this letter, we present a 3-dB power divider with isolated port. The input port is a rectangular waveguide and two output ports are microstrip lines. The measured results of the power divider show above 12-dB minimum isolation over 13-16 GHz, with an input return loss of better than 15 dB. The return loss of the output ports is better than 10 dB, with an insertion loss of lower than 0.35 dB. A maximum amplitude imbalance of about 0.15 dB is observed within 13-16 GHz, with a phase imbalance of better than 2.5°. The proposed power divider in this letter is compact and good integration with the monolithic microwave integrated circuit (MMIC) devices.

BACKGROUND: The dose of radiation a patient receives when undergoing dual-energy computed tomography (CT) is of significant concern to the medical community, and balancing the tradeoffs between the level of radiation used and the quality of CT images is challenging. This paper proposes a method of synthesizing high-energy CT (HECT) images from low-energy CT (LECT) images using a neural network that achieves an alternative to HECT scanning by employing an LECT scan, which greatly reduces the radiation dose a patient receives. METHODS: In the training phase, the proposed structure cyclically generates HECT and LECT images to improve the accuracy of extracting edge and texture features. Specifically, we combine multiple connection methods with channel attention (CA) and pixel attention (PA) mechanisms to improve the network's mapping ability of image features. In the prediction phase, we use a model consisting of only the network component that synthesizes HECT images from LECT images. RESULTS: Our proposed method was conducted on clinical hip CT image data sets from Guizhou Provincial People’s Hospital. In a comparison with other available methods [a generative adversarial network (GAN), a residual encoder-to-decoder network with a visual geometry group (VGG) pretrained model (RED-VGG), a Wasserstein GAN (WGAN), and CycleGAN] in terms of metrics of peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM), normalized mean square error (NMSE), and a visual effect evaluation, the proposed method was found to perform better on each of these evaluation criteria. Compared with the results produced by CycleGAN, the proposed method improved the PSNR by 2.44%, the SSIM by 1.71%, and the NMSE by 15.2%. Furthermore, the differences in the statistical indicators are statistically significant, proving the strength of the proposed method. CONCLUSIONS: The proposed method synthesizes high-energy CT images from low-energy CT images, which significantly reduces both the cost of treatment and the radiation dose received by patients. Based on both image quality score metrics and visual effects comparisons, the results of the proposed method are superior to those obtained by other methods.

BACKGROUND: We aimed to compare the performance of three contrast-enhanced T1-weighted three-dimensional (3D) magnetic resonance (MR) sequences to detect brain tumors at 3 Tesla. The three sequences were: (I) delay alternating with nutation for tailored excitation sampling perfection with application-optimized contrasts using different flip angle evolution (DANTE-SPACE), (II) pointwise encoding time reduction with radial acquisition (PETRA), and (III) magnetization-prepared rapid acquisition with gradient echo (MPRAGE). METHODS: This study involved 77 consecutive patients, including 34 patients with known primary brain tumors and 43 patients suspected of intracranial metastases. All patients underwent each of the three sequences with comparable spatial resolution and acquisition time post-injection. Signal-to-noise ratios (SNRs) for gray matter (GM) and white matter (WM), contrast-to-noise ratios (CNRs) for lesion/GM, lesion/WM, and GM/WM were quantitatively compared. Two radiologists determined the total number of enhancing lesions by consensus. Intraclass correlation coefficients (ICCs) between the two radiologists for metastases presence, qualitative ratings for image quality, and acoustic noise level of each sequence were assessed. RESULTS: Among the three sequences, SNRs and CNRs between lesions and surrounding parenchyma were highest using DANTE-SPACE, but CNR(WM/GM) was the lowest with DANTE-SPACE. SNRs for PETRA images were significantly higher than those for MPRAGE (P0.05). Significantly more brain metastases were detected with DANTE-SPACE (n=94) compared with MPRAGE (n=71) and PETRA (n=72). The ICCs were 0.964 for MPRAGE, 0.975 for PETRA, and 0.973 for DANTE-SPACE. Qualitative scores for lesion imaging using DANTE-SPACE were significantly higher than those obtained with PETRA and MPRAGE (P=0.002 and P=0.004, respectively). The acoustic noise level for PETRA (64.45 dB) was significantly lower than that for MPRAGE (78.27 dB, P

In this study, a basic organic Rankine cycle (ORC) is introduced in an air separation process for waste heat recovery. Conventional and advanced exergy analyses are adopted to investigate the thermodynamic properties of components in the ORC. A comprehensive thermodynamic model is constructed to improve the advanced exergy analysis in the ORC, thereby encompassing real, theoretical, unavoidable, and hybrid cycles. Nine organic working fluids are introduced to investigate the influence on the ORC performance. (1) The conventional exergy analysis reveals the following: (a) The expander constantly demonstrates the maximum exergy efficiency except when R227ea is used. (b) The evaporator constantly exhibits the maximum exergy destruction regardless of the working fluid used. (c) The maximum product exergy is obtained when R114 is used. (d) Key components must focus on the condenser and evaporator to improve the ORC performance. (2) The advanced exergy analysis reveals that the expander demonstrates maximum potential for improvement because its endogenous avoidable exergy destruction accounts for approximately 90% of its real exergy destruction for all working fluids. The expander must be improved to achieve the optimal ORC performance. The advanced exergy analysis can distinguish the source of exergy destruction and the magnitude for possible improvement via the proposed thermodynamic model in this study. The comprehensive thermodynamic model can promote the investigation of the advanced exergy analysis in the ORC. Applying conventional and advanced exergy analyses to investigate the thermodynamic performance of a system or its components is highly recommended.

Background Persistent infection with high-risk types Human Papillomavirus could cause diseases including cervical cancers and oropharyngeal cancers. Nonetheless, so far there is no effective pharmacotherapy for treating the infection from high-risk HPV types, and hence it remains to be a severe threat to health of female. Methods In light of drug repositioning strategy, we trained and benchmarked multiple machine learning predictive models so as to predict potential effective antiviral drugs for HPV infection in this work. Based on antiviral-target interaction dataset, we generated high dimension feature set of drug-target interaction pairs and used the dataset to train and construct machine learning predictive models. Results Through optimizing models, measuring models’ predictive performance using 182 pairs of antiviral-target interaction dataset which were all approved by United States Food and Drug Administration, and benchmarking different models’ predictive performance, we identified the optimized Support Vector Machine and K-Nearest Neighbor classifier with high precision score were the best two predictors (0.80 and 0.85 respectively) amongst classifiers of Support Vector Machine, Random forest, Adaboost, Naive Bayes, K-Nearest Neighbors, and Logistic regression classifier. We applied these two predictors together and successfully predicted 58 pairs of antiviral-HPV protein interactions from 846 pairs of antiviral-HPV protein associations. Conclusions Our work provided good drug candidates for anti-HPV drug discovery. So far as we know, we are the first one to conduct such HPV-oriented computational drug repositioning study.

BACKGROUND The treatment of small-cell lung cancer (SCLC) has progressed little in recent years because of its unique biological activities and complex genomic alterations. Chemotherapy combined with radiotherapy has been widely accepted as the first-line treatment for SCLC. CASE SUMMARY Here, we present a 68-year-old male smoker who was diagnosed with SCLC of the right lung. After several cycles of concurrent chemoradiotherapy, the tumor progressed with broad metastasis to liver and bone. Histopathological examination showed an obvious transformation to adenocarcinoma, probably a partial recurrence mediated by the chemotherapy-based regimen. A mixed tumor as the primary lesion and transformation from SCLC or/and tumor stem cells may have accounted for the pathology conversion. We adjusted the treatment schedule in accord with the change in phenotype. CONCLUSION Although diffuse skeletal and hepatic metastases were seen on a recent computed tomography scan, the patient is alive, with intervals of progression and shrinkage of his cancer.

Carbon nanotube field-effect transistor (CNTFET) shows great potential for digital and analog applications due to the unique 1-D ballistic carrier transport in carbon nanotubes (CNTs). A new compact physical model for CNT Schottky-barrier transistors is proposed in this article. The Schottky barrier between the contact metal and CNTs is described by using an effective barrier height, and more importantly, the influence of using CNT arrays or networks rather than a single CNT as the device channel is considered in this model. Moreover, the trapping effect is investigated and modeled by adding RC delay element to the equivalent circuit. The model is proposed in a concise form to avoid complex operations such as integration and is compatible with circuit simulation. The simulation results of I-V characteristics are verified to be in agreement with experimental data.

Summary Surface energy of the underneath interlayer plays a crucial role in determining the morphology of bulk-heterojunction (BHJ) film in solution-processed organic solar cells (OSCs). However, little attention has been paid to this perspective at the microscopic level. In this study, we demonstrate an effective strategy to tune the nanoscale heterogeneous distribution of surface energy (HeD-SE) at interlayers through incorporating two-dimensional (2D) MoS2 nanosheets, which modulate the component distribution, molecular orientation, and phase separation in the upper BHJ layers, leading to suppressed charge recombination and increased charge extraction in devices with enhanced power conversion efficiency (PCE) and device stability. Furthermore, we explore the relationship among the different surface energies of BHJs between the donor and the acceptor, the nanoscale HeD-SE at interlayers, and the PCEs of the resultant devices. This work reveals the impact of nanoscale heterogeneity in the distribution of surface energy in the underneath interlayers on the BHJ morphology and device performance in OSCs.

Electric discharge machining (EDM) can be used to cut complex parts and hard materials, making it one of the most popular machining processes today. However, the energy efficiency of EDM is low compared with conventional methods such as turning, milling and grinding. In addition, the gas and noise emissions of the EDM process affect the environment and the health of operators. This paper reviews the research trends in the latest sustainable techniques in EDM. First, a detailed review of existing EDM techniques for enhancing energy efficiency conducted. The review considers three aspects: energy distribution, the relationship between process parameters and specific discharge energy, and the effects of external field assistance. Second, the dry-EDM process-a clean manufacturing technique is introduced and its environmental impact is discussed. The machining performance indicators of dry-EDM, such as the material removal rate, tool wear ratio, surface topography, and microstructures, are compared with those of conventional EDM. Next, the indicators of gas and noise emissions in EDM are proposed to evaluate the sustainable manufacturing process. The relationship between the process parameters and emission indicators is reviewed, and multi-optimization considering machining noise is discussed. Finally, the impact of sustainable techniques in EDM on the environment, society, and economy are discussed, as well as the application costs of these techniques. Some directions for future research are also proposed.

Purpose 68Ga-PSMA PET/CT has high specificity and sensitivity for the detection of both intraprostatic tumor focal lesions and metastasis. However, approximately 10% of primary prostate cancer are invisible on PSMA-PET (exhibit no or minimal uptake). In this work, we investigated whether machine learning-based radiomics models derived from PSMA-PET images could predict invisible intraprostatic lesions on 68Ga-PSMA-11 PET in patients with primary prostate cancer.Methods In this retrospective study, patients with or without prostate cancer who underwent 68Ga-PSMA PET/CT and presented negative on PSMA-PET image at either of two different institutions were included: institution 1 (between 2017 to 2020) for the training set and institution 2 (between 2019 to 2020) for the external test set. Three random forest (RF) models were built using selected features extract from standard PET images, delayed PET images, and both standard and delayed PET images. Then, subsequent 10-fold cross-validation was performed. In the test phase, the three RF models and PSA density (PSAD, cut-off value: 0.15ng/ml/ml) were tested with the external test set. The area under the receiver operating characteristic curve (AUC) was calculated for the models and PSAD. The AUCs of the radiomics model and PSAD were compared.Results A total of 64 patients (39 with prostate cancer and 25 with benign prostate disease) were in the training set, and 36 (21 with prostate cancer and 15 with benign prostate disease) were in the test set. The average AUCs of the three RF models from 10-fold cross-validation were 0.87 (95% CI: 0.72, 1.00), 0.86 (95% CI: 0.63, 1.00) and 0.91 (95% CI: 0.69, 1.00), respectively. In the test set, the AUCs of the three trained RF models and PSAD were 0.903 (95% CI: 0.830, 0.975), 0.856 (95% CI: 0.748, 0.964), 0.925 (95% CI:0.838, 1.00), and 0.662 (95% CI: 0.510, 0.813). The AUCs of the three radiomics models were higher than that of PSAD (0.903, 0.856 and 0.925 vs 0.662, respectively; P = .007, P = .045 and P = .005, respectively).Conclusion Random forest models developed by 68Ga-PSMA-11 PET-based radiomics features were proven useful for accurate prediction of invisible intraprostatic lesion on 68Ga-PSMA-11 PET in patients with primary prostate cancer and showed better diagnostic performance compared with PSAD.

The design of photocatalysts with hierarchical pore sizes is an effective method to improve mass transport, enhance light absorption, and increase specific surface area. Moreover, the construction of a heterojunction at the interface of two semiconductor photocatalysts with suitable band positions plays a crucial role in separating and transporting charge carriers. Herein, ZIF-8 and urea are used as precursors to prepare hierarchically porous ZnO/g-C3N4 S-scheme heterojunction photocatalysts through a two-step calcination method. This S-scheme heterojunction photocatalyst shows high activity toward photocatalytic H2O2 production, which is 3.4 and 5.0 times higher than that of pure g-C3N4 and ZnO, respectively. The mechanism of charge transfer and separation within the S-scheme heterojunction is studied by Kelvin probe, in situ irradiated X-ray photoelectron spectroscopy (ISI-XPS), and electron paramagnetic resonance (EPR). This research provides an idea of designing S-scheme heterojunction photocatalysts with hierarchical pores in efficient photocatalytic hydrogen peroxide production.

The detection of harmful trace gases, such as formaldehyde (HCHO), is a technical challenge in the current gas sensor field. The weak electrical signal caused by trace amounts of gases is difficult to be detected and susceptible to other gases. Based on the amplification effect of a field-effect transistor (FET), a carbon-based FET-type gas sensor with a gas-sensing gate is proposed for HCHO detection at the ppb level. Semiconducting carbon nanotubes (s-CNTs) and a catalytic metal are chosen as channel and gate materials, respectively, for the FET-type gas sensor, which makes full use of the respective advantages of the channel transport layer and the sensitive gate layer. The as-prepared carbon-based FET-type gas sensor exhibits a low detection limit toward HCHO up to 20 ppb under room temperature (RT), which can be improved to 10 ppb by a further heating strategy. It also exhibits a remarkable elevated recovery rate from 80 to 97% with almost no baseline drift (2%) compared to the RT condition, revealing excellent reproducibility, stability, and recovery. The role of sensitive function in the FET-type gas sensor is performed by means of an independent gas-sensing gate, that is, the independence of the sensitive gate and the electron transmission channel is the main reason for its high sensitivity detection. We hope our work can provide an instructive approach for designing high-performance formaldehyde sensor chips with on-chip integration potential.

Spectroscopic profiling data (e.g., Raman spectroscopy and mass spectroscopy), combined with machine learning, have provided a data-driven approach for discriminative tasks. In these tasks, researchers often start with simple classification models. If one model doesn’t work, they will try more sophisticated models. If all models fail, the researchers will deem the data set as “inseparable.“ This “trial-and-error” practice reveals a fundamental question: does the dataset possess the necessary statistical power for the current discriminative task? This “classifiability analysis” is an implicit and often neglected step in the data-driven pipeline. This paper aims to design a unified methodological framework for classifiability analysis. In this framework, a meta-learner model combines diversified atom metrics (e.g., Bayes error rate / irreducible error, classification accuracy, information gain / mutual information) into one unified metric (d). We have successfully used the proposed framework to analyze a spectroscopic profiling dataset to discriminate vintage liquors of different ages. A significant difference (d = 1.447. d > 0.8 indicates a significant difference) between 5-year and 16-year liquors.

Increasing evidence reveals that physical exercise is an efficient therapeutical approach in the treatment of insulin resistance (IR) and related metabolic diseases. However, the potential beneficial effects of exercise on insulin resistance and its underlying mechanisms remain unclear. Recent findings elucidated the negative role of ASK1 in repressing the glucose uptake through JNK1‐IRS1‐Akt signalling in liver. Thus, a detailed investigation of the effect of ASK1‐mediated insulin signalling on exercise‐mediated improvement of insulin sensitivity and its underlying mechanism was implemented in this study. Using a high‐fat diet‐induced IR rat model of chronic or acute swimming exercise training, we here showed that body weight and visceral fat mass were significantly reduced after chronic exercise. Moreover, chronic exercise reduced serum FFAs levels and hepatic triglyceride content. Both chronic and acute exercise promoted glucose tolerance and insulin sensitivity. Meanwhile, both chronic and acute exercise decreased ASK1 phosphorylation and improved JNK1‐IRS1‐Akt signalling. Furthermore, exercise training decreased CFLAR, CREG and TRAF1 protein levels in liver of obese rats, which are positive regulator of ASK1 activity. These results suggested that swimming exercise demonstrated to be an effective ameliorator of IR through the regulation of ASK1‐mediated insulin signalling and therefore, could present a prospective therapeutic mean towards the treatment of IR and several metabolic diseases based on IR, containing NAFLD and type Ⅱ diabetes.

Objective Helicobacter pyloriinfection is mostly a family-based infectious disease. To facilitate its prevention and management, a national consensus meeting was held to review current evidence and propose strategies for population-wide and family-basedH. pyloriinfection control and management to reduce the related disease burden. Methods Fifty-seven experts from 41 major universities and institutions in 20 provinces/regions of mainland China were invited to review evidence and modify statements using Delphi process and grading of recommendations assessment, development and evaluation system. The consensus level was defined as ≥80% for agreement on the proposed statements. Results Experts discussed and modified the original 23 statements on family-basedH. pyloriinfection transmission, control and management, and reached consensus on 16 statements. The final report consists of three parts: (1)H. pyloriinfection and transmission among family members, (2) prevention and management ofH. pyloriinfection in children and elderly people within households, and (3) strategies for prevention and management ofH. pyloriinfection for family members. In addition to the ‘test-and-treat’ and ‘screen-and-treat’ strategies, this consensus also introduced a novel third ‘family-basedH. pyloriinfection control and management’ strategy to prevent its intrafamilial transmission and development of related diseases. Conclusion H. pyloriis transmissible from person to person, and among family members. A family-basedH. pyloriprevention and eradication strategy would be a suitable approach to prevent its intra-familial transmission and related diseases. The notion and practice would be beneficial not only for Chinese residents but also valuable as a reference for other highly infected areas.

With the development of Internet of Things technology, Channel State Information (CSI) based human activity recognition (HAR) plays an important role in Human-Computer Interaction and achieves considerable advancements over recent years. However, when the trained model is applied to recognize new activity categories or recognizing new users in new scenarios, the recognition performance of general methods will dramatically decline. And re-collecting adequate new activity categories’ samples to train the HAR model to adapt to the new situation will consume a lot of time and human effort. To overcome this challenge, we propose a framework, Augment Few Shot Learning-based Human Activity Recognition (AFSL-HAR), which can achieve significant performance in recognizing new categories through a small amount of samples to fine-tune the model parameters and avoid retraining the network from scratch again. And besides, in order to improve the robustness of AFSL-HAR, we design a Feature Wasserstein Generative Adversarial Network (FWGAN) module, which can synthesize diverse samples to help the recognition model learn more sharper classification boundaries. Specifically, the FWGAN module incorporates a feature extractor to realize converging with a fewer number of training samples, and takes an improved discriminator to enhance system performance. The experimental results demonstrate that AFSL-HAR can achieve accuracy of 98.9% and 94.7% when recognizing new activities using few samples for each category on the public data set and self-made data set, respectively.

This paper studies the impacts of money-back guarantees (MBGs) in the presence of parallel importation, where a manufacturer sells a product facing customer fit uncertainties in two markets and a parallel importer diverts unauthorized products from the low-price market to the high-price one. The two firms compete in both pricing and MBG strategies. We consider two types of parallel importers, a third-party agent and an authorized retailer in the low-price market, respectively. By applying the certainty equivalent pricing approach, we show that offering MBGs only changes firms’ marginal costs without affecting customers’ purchasing decisions directly. When the net salvage values of returned products are positive in the high-price market, both the manufacturer and the third-party agent offer MBGs. Whenever the manufacturer offers an MBG in the low-price market, there is a price inflation effect that counters parallel importation. Therefore the manufacturer may offer the MBG even if the net salvage value is negative. When the authorized retailer decides refund policies in two markets, the MBG strategy in one market changes the retailer’s profit from the authorized channel and that from the gray market in opposite directions, and has an interaction with the refund policy in another market. No refund may be preferred in either market even if the net salvage value is positive. We find that the manufacturer’s MBG strategy in the high-price market always deters parallel importation while other MBG strategies may achieve Pareto improvements in the equilibrium.

Glycoproteomics is a powerful yet analytically challenging research tool. Software packages aiding the interpretation of complex glycopeptide tandem mass spectra have appeared, but their relative performance remains untested. Conducted through the HUPO Human Glycoproteomics Initiative, this community study, comprising both developers and users of glycoproteomics software, evaluates solutions for system-wide glycopeptide analysis. The same mass spectrometrybased glycoproteomics datasets from human serum were shared with participants and the relative team performance for N - and O-glycopeptide data analysis was comprehensively established by orthogonal performance tests. Although the results were variable, several high-performance glycoproteomics informatics strategies were identified. Deep analysis of the data revealed key performance-associated search parameters and led to recommendations for improved ‘high-coverage’ and ‘high-accuracy’ glycoproteomics search solutions. This study concludes that diverse software packages for comprehensive glycopeptide data analysis exist, points to several high-performance search strategies and specifies key variables that will guide future software developments and assist informatics decision-making in glycoproteomics., This analysis presents the results of a community-based evaluation of existing software for large-scale glycopeptide data analysis.

Shan-Shan Cui,1 Li-Jun Duan,1 Jun-Feng Li,1 Yong-Zhang Qin,2 Su-Qing Bao,1 Xia Jiang1 1Department of Endocrinology and Metabolism, Tianjin First Central Hospital, Tianjin, People’s Republic of China; 2Department of Endocrinology, First Affiliated Hospital of Gannan Medical University, Ganzhou, People’s Republic of ChinaCorrespondence: Shan-Shan CuiDepartment of Endocrinology and Metabolism, Tianjin First Central Hospital, No. 24 of Fukang Road, Nankai District, Tianjin, People’s Republic of ChinaTel +86 15022353629Email shanshan87@tom.comObjective: This study aims to explore the factors influencing the renal glucose threshold (RTG) in patients with newly diagnosed type 2 diabetes mellitus (T2DM).Methods: A cross-sectional study was conducted on 1009 hospitalized patients with T2DM using stratified random sampling. Blood glucose was monitored using a dynamic blood glucose monitor to obtain the mean blood glucose (MBG), which is used to calculate the RTG. The factors influencing the RTG were then analyzed.Results: The mean RTG in patients with newly diagnosed T2DM was 203.58 ± 55.22 mg/dl. The correlation between the RTG and the various variables was analyzed, and the results demonstrated that the RTG was correlated with the patient’s age (r = − 0.14539, P = 0.0001); MBG (r = − 0.35009, P = 0.0001); renal long neck (r = 0.16762, P = 0.0001); homeostatic model assessment for insulin resistance (r = − 0.38322, P = 0.0001); homeostatic model assessment for beta-cell function (r = − 0.22770, P = 0.0001); and the levels of glycated hemoglobin (HbA1c; r = 0.98994, P = 0.0001), blood urea nitrogen (r = − 0.11093, P = 0.0004), creatinine (r = − 0.26414, P = 0.0001), uric acid (r = − 0.20149, P = 0.0001), total cholesterol (r = 0.13192, P = 0.0001), low-density lipoprotein (r = 0.12466, P = 0.0001), thyroid-stimulating hormone (r = − 0.06346, P = 0.0460), beta-2 microglobulin (r = − 0.08884, P = 0.0056), and 24-hour urine glucose (r = 0.32115, P = 0.0001). Multiple linear stepwise regression analysis revealed that the HbA1c, 24-hour urine glucose, estimated glomerular filtration rate (eGFR), D-dimer, and body mass index (BMI) should be included in the final model, and HbA1c had the greatest impact on the RTG followed in descending order by the 24-hour urine glucose, eGFR, D-dimer, and BMI (P < 0.05).Conclusion: The RTG increases in most patients with newly diagnosed diabetes. The risk factors for the RTG are HbA1c, 24-hour urine glucose, eGFR, D-dimer, and BMI.Keywords: newly diagnosed, type 2 diabetes mellitus, renal glucose threshold, continuous glucose monitoring

Background/Aim: This study mainly aimed to explore the therapeutic effects of 3 renal replacement therapy (RRT) modalities on acute kidney injury (AKI) caused by wasp stings. Methods: A retrospective study from September 2016 to December 2019 was conducted. Thirty-one patients with AKIs caused by wasp sting were selected and divided into 3 groups according to the initial RRT modality received, namely, (1) the intermittent hemodialysis combined with hemoperfusion (IHD + HP) group, (2) the continuous veno-venous hemodiafiltration (CVVHDF) group, and (3) the CVVHDF combined with HP (CVVHDF + HP) group. The laboratory results were measured and analyzed before treatment on the 3rd, 7th, and 14th days of treatment. The renal function outcomes and survival of the patients were investigated at 3 months follow-up. Results: The laboratory results of enzyme measures and inflammatory indicators in wasp sting patients increased significantly in the early stage and 3 RRT modalities were effective in reducing these indicators. In addition, continuous RRT modality (CVVHDF and CVVHDF + HP) showed better clearance of myoglobin than IHD + HP. The serum creatinine levels of patients in the 3 groups did not recover to baseline within 14 days after beginning treatment. Nevertheless, the CVVHDF + HP group was better than the CVVHDF group, and CVVHDF was better than the IHD + HP group on the 3rd day. The interleukin (IL)-6 and IL-10 levels in CVVHDF + HP and IHD + HP groups were obviously lower than those in the CVVHDF group on the 3rd day. In the follow-up study, the recovery rate of renal function in CVVHDF and CVVHDF + HP groups was significantly better than that in the IHD + HP group. Conclusion: Early RRT was effective in the treatment of patients with A KI caused by wasp sting. CVVHDF + HP and CVVHDF modalities were better than the IHD + HP group in venom clearance and renal function recovery.

Purpose To investigate retinal nerve fiber layer (RNFL) thickness changes at different stages of Leber's hereditary optic neuropathy (LHON) in patients bearing the ND4 G11778A mutation. Methods Ninety-eight clinically diagnosed, G11778A-positive LHON patients underwent 538 optical coherence tomography (OCT) examinations from September 2015 to September 2017. Patients were grouped based on disease duration at examination. Fifty healthy volunteers underwent 100 examinations as controls. Differences in RNFL thickness were compared across groups. Results During the onset of LHON patients with G11778A mutation, the thickness of nerve fiber layer in temporal quadrant decreased slowly within 1-3 months (p > .05), then entered in the rapid thinning period, which generally lasted until about 12 months of the course of disease (p .05); The optic nerve fiber layer in other quadrants was usually stayed in a significant thickening period within 1-3 months (p .05) . Conclusion In LHON patients with G11778A mutation, the thickness of optic nerve fiber layer in the temporal side will experience slow thinning stage, rapid thinning stage and stable stage; The thickness of optic nerve fiber layer in other directions varies with the course of disease. Generally, it will experience five periods: significant thickening period, swelling period, recovery period, rapid thinning period and stable period.

BACKGROUND While tuberculosis (TB) itself is a common disease, isolated TB of the liver is a rare entity. Tubercular involvement of the liver is more commonly a part of a disseminated disease of the hepatic parenchyma. In contrast, isolated hepatic TB spread through the portal vein from the gastrointestinal tract is seldom encountered in clinical practice, with only a few sporadic cases and short series available in the current literature. Vascular complications, such as portal vein thrombosis (PVT), have rarely been reported previously. CASE SUMMARY A 22-year-old man was hospitalized with complaints of a 3-mo history of fever and weight loss of approximately 10 kg. He had a 10-year hepatitis B virus (HBV) infection in his medical history. Contrast-enhanced computed tomography (CECT) confirmed hepatosplenomegaly, with hypodensity of the right lobe of the liver and 2.1 cm thrombosis of the right branch of the portal vein. A liver biopsy showed epithelioid granulomas with a background of caseating necrosis. Ziehl-Nelson staining showed acid-fast bacilli within the granulomas. The patient was diagnosed with isolated hepatic TB with PVT. Anti-TB therapy (ATT), including isoniazid, rifapentine, ethambutol, and pyrazinamide, was administered. Along with ATT, the patient was treated with entecavir as an antiviral medication against HBV and dabigatran as an anticoagulant. He remained asymptomatic, and follow-up sonography of the abdomen at 4 mo showed complete resolution of the PVT. CONCLUSION Upon diagnosis of hepatic TB associated with PVT and HBV coinfection, ATT and anticoagulants should be initiated to prevent subsequent portal hypertension. Antiviral therapy against HBV should also be administered to prevent severe hepatic injury.

Lithium-ion hybrid capacitors (LIHCs), as a novel energy storage device combining the mechanism of lithium-ion batteries and supercapacitors, can achieve high energy and power density simultaneousl...

To evaluate the efficacy of acupuncture combined with rehabilitation training in patients with intensive care unit (ICU)-acquired muscle weakness (ICUAW), a single-blinded, randomized, sham-controlled clinical trial is designed for execution. In total, 56 participants with ICUAW will be randomly assigned to the treatment and control groups with 28 participants in each group. The participants will be treated with acupunctures or sham procedures at LI15, LI11, ST36, GB34, and ST31, 5 times per week for a total of 20 sessions in 4 weeks while they will receive rehabilitation training. Patients will be followed up every month for 3 months after treatment. The primary outcomes include changes in quadriceps femoris muscle area, thickness, vastus intermediate muscle thickness, subcutaneous tissue thickness, and ultrasonic intensities of the rectus femoris. The secondary outcomes consist of the modified Barthel index score and the Medical Research Council total score. Participants’ mechanical ventilation, the rate of detachment at the second week, the 28-day survival rate, and the occurrence of adverse reactions will be measured, and any side effects will be reported and recorded. Patient outcomes between the treatment and control groups will be compared and statistically tested. We anticipate that the therapeutic regimen of acupuncture combined with rehabilitation training would be more effective than the rehabilitation training alone for the treatment of the ICUAW. The findings of this study could help develop a better strategy for the treatment of the ICUAW disease and explore a clinical application of an acupuncture technique. Trial registration: Chinese Clinical Trial Register ChiCTR2000038779. Registered 30 September, 2020, https://www.chictr.org.cn/showproj.aspx?proj=62284.

Lead halide perovskite quantum dots (PQDs) exhibit excellent photoelectric and optical properties, but their poor stability and low multiphoton absorption efficiency greatly limit their biological applications. Efforts have been made to combine upconversion nanoparticles (UCNPs) with PQDs to produce a composite material that is NIR-excitable, upconverting, and emission-tunable due to the unique optical properties of UCNPs, which converts tissue-penetrating near-infrared light into visible light based on an upconversion multiphoton excitation process. However, it is challenging to make such a nanocrystal heterostructure and maintain good optical properties and stability of both UCNPs and PQDs because they have different crystal structures. Here, we report the synthesis of heterostructured UCNP-PQD nanocrystals to bring hexagonal-phase NaYF4 UCNPs and cubic-phase CsPbBr1X2 PQDs in close proximity in a single nanocrystal, leading to efficient Forster resonance energy transfer (FRET) from the UCNP to the PQD under NIR excitation, as compared to their counterparts in solution. Moreover, by further improving the lattice matching between the UCNP and PQD using Gd to replace Y, heterostructured CsPbBr3-NaGdF4:Yb,Tm nanocrystals are successfully synthesized, with much enhanced luminescence and stability at high temperatures or in polar solvents or under continuous ultraviolet light excitation as compared to those of the CsPbBr3-NaYF4:Yb,Tm nanocrystals and pure PQDs.

Blood cysts are rare cardiac tumors, often involve the atrioventricular valves. Blood cysts are usually small and tend to be clinically silent. Here, we report a case of blood cyst adhered to the junction of the pulmonary valve and the right ventricular outflow, causing severe outflow obstruction. This case was initially misdiagnosed as pulmonary valve stenosis by transthoracic echocardiography (TTE) and even underwent percutaneous balloon pulmonary valvuloplasty (PBPV). As the right ventricle-to-pulmonary artery peak gradient did not decrease after PBPV procedure, the boy underwent TTE again and the diagnosis was corrected as blood cyst. The cystic mass was successfully resected subsequently, and the boy recovered well.

In the past decades, Dynamic Internet of Things (D-IoT) networks have played a conspicuously more important role in many real-life areas, including disaster relief, environment monitoring, public safety, and so on, to rapidly collect information from the environment and help people to make the decision. Meanwhile, due to the widespread implementation of dynamic IoT networks, there exists an enormous demand on designing suitable models and efficient algorithms for fundamental operations in dynamic IoT networks, to achieve the high throughput and reliable low-latency communication demands in 6G networks. In this article, we first present a general dynamic model to comprehensively depict most of the dynamic phenomena in IoT networks. Then, based on the proposed dynamic model, a distributed scheduling algorithm is proposed to competitively optimize the Age-of-Information (AoI) problem in the context of a D-IoT network. We say our scheduling algorithm is competitive: the throughput of the base station approximates the optimal solution with constant competitive ratio; and, the latency for a packet received by the base station is only constant times larger than the optimal latency. Rigorous theoretical analysis and extensive simulations are presented to verify the high throughput and reliable low-latency communications in our proposed algorithm.

As the size of dataset increases, it has become a difficult task to explore structures of interest from crowded parallel coordinates due to visual clutter. Numerous methods have been proposed to simplify the visualization of crowded parallel coordinates, such as filtering, bundling and sampling. However, contextual structures are hardly preserved in the course of simplification, which make significant features easily lost in the simplified parallel coordinates. In this paper, we propose a context-aware visual sampling method for the exploration of crowded parallel coordinates. A Doc2Vec model, widely used in the field of Natural Language Processing (NLP), is utilized to represent the contextual structures across a series of attribute axes with quantifiable vectors. Then, an adaptive blue noise sampling model is employed to reduce the size of original dataset in the vectorized space, guarantying that data items with different contextual structures would be retained in the simplified parallel coordinates. A set of meaningful visual interfaces are designed, enabling users to easily capture the contextual features and evaluate the sampled parallel coordinates. Case studies based on real-world datasets and quantitative comparison have demonstrated the effectiveness of our method in the simplification of crowded parallel coordinates and the exploration of large scale multi-dimensional datasets.

The successful applications of two-dimensional (2D) transition metal dichalcogenides highly rely on rational regulation of their electronic properties. The nondestructive and controllable doping strategy is of great importance to implement 2D materials in electronic devices. Herein, we propose a straightforward and effective method to realize controllable n-type doping in WSe2 monolayer by electron beam irradiation. Electrical measurements and photoluminescence (PL) spectra verify the strong n-doping in electron beam-treated WSe2 monolayers. The n-type doping arises from the generation of Se vacancies and the doping degree is precisely controlled by irradiation fluences. Due to the n-doping-induced narrowing of the Schottky barrier, the current of back-gated monolayer WSe2 is enhanced by an order of magnitude and a ∼8× increase in the electron filed-effect mobility is observed. Remarkably, it is a moderate method without significant reduction in electrical performance and severe damage to lattice structures even under ultra-high doses of irradiation.

Objective To analyze the rates of blindness with the demographics and clinical characteristics of patients with primary angle-closure disease (PACD) to provide a comprehensive epidemiologic reference in China. Methods A retrospective analysis was conducted in the Chinese Glaucoma Study Consortium database, which is a national multicenter glaucoma research alliance of 111 hospitals participating between December 21, 2015 and September 9, 2018. The diagnosis of PACD was made by qualified physicians through examination. Comparison of sex, age, family history, subtypes of PACD, and blindness were analyzed. Results A total of 5762 glaucoma patients were included, of which 4588 (79.6%) had PACD. Of PACD patients, 72.1% were female with the sex ratio (F/M) of 2.6, and the average age of patients was 63.8±9.3 years with the majority between 60 and 70 years. Additionally, 30% of these patients had low vision in one eye, 8.8% had low vision in both eyes, 1.7% had blindness in one eye, and 0.3% had blindness in both eyes. There were statistical differences with regards to age between male and female patients with PACD, with male patients being older on average. Primary angle-closure glaucoma was more commonly diagnosed in males (60%) compared to females (35.9%), whereas acute primary angle closure (APAC) was more commonly diagnosed in females (54.3%) compared to males (37.7%). The visual acuity in APAC patients was lower and the rate of low vision and blindness was higher than other subtypes. Conclusion PACD was the major type of glaucoma in Chinese hospitals. There were more female patients with PACD, mostly between 60 and 70 years old, with higher rates of APAC in women. APAC resulted in the worst visual outcomes of all PACD subtypes.

BACKGROUND: Prostate cancer is one of the most common malignant tumors worldwide, and is the third-leading cause of cancer death in men. Nearly 70% of new prostate cancer patients in China are locally advanced or widely metastatic with poor prognosis. Providing active treatment to early stage prostate cancer patients can improve the prognosis of prostate cancer patients. Thus, this study sought to evaluate the economy of early prostate specific antigen (PSA) screening for high-risk prostate cancer. METHODS: Based on the data collected from the PSA screening activities of 11 county hospitals from October 2019 to April 2021, this study evaluated a high-risk prostate cancer population who received PSA screening and their quality of life and economy. The screening population comprised males aged over 50 years. All screening patients were tested for PSA. If the PSA value is unnormal, a further diagnosis based on magnetic resonance imagining (MRI) or a transrectal ultrasound-guided prostate biopsy were performed. The decision-tree and Markov model was used to simulate the process of disease development of high-risk prostate cancer patients who underwent screenings and those who did not, and the incremental cost-effectiveness ratio was also evaluated. RESULTS: A total of 13,726 men received a PSA screening. Of these, 1,062 men had abnormal PSA values, and 73 of these were diagnosed with prostate cancer. Of these 73 patients, 40, 21, and 12 had early stage, mid-stage, and late-stage prostate cancer, respectively. Compared to the patients unscreened, the 1,000 patients who received an early PSA screening increased their quality-adjusted life year (QALY) by 15.69 years; however, each QALY had an additional cost of 38,550 yuan, which was lower than the willingness to pay threshold of 72,447 yuan (per capita gross domestic product in 2020). CONCLUSIONS: For high-risk prostate cancer patients, early screenings have a cost-effective advantage over no screenings. Thus, early screening should be vigorously promoted for high-risk prostate cancer patients.

Background: This study explored the clinical role of lncRNA MEG3 in rheumatoid arthritis (RA) management. Materials & methods: Totally, 191 active RA patients were enrolled, and their lncRNA MEG3 expressions in peripheral blood monoclonal cells were detected. Results: LncRNA MEG3 expression was downregulated, and it negatively correlated with lesion joints, inflammation and disease activity in RA patients. Moreover, lncRNA MEG3 expression was increased during treatment; meanwhile its increment correlated with treatment response and remission. Conclusion: LncRNA MEG3 may serve as a potential biomarker for monitoring treatment efficacy in RA management.

Background Synovial sarcoma (SS) is a highly malignant tumor of unknown histological origin. This tumor can occur in various parts of the body, including those without synovial structures, but mainly in and around the joints, mostly in the lower extremities. Primary intracranial SSs are remarkably rare. This paper aims to report a case of primary intracranial SS with hemorrhage. Case summary A 35-year-old male patient suffered a headache and slurred speech during manual labor and was sent to the emergency department. Through imaging examination, the patient was considered to have high-grade glioma complicated with hemorrhage and was treated with craniotomy. Postoperative pathology revealed SS. positron emission tomography/computed tomography was performed, which ruled out the possibility of metastasis to the intracranial from other parts of the body. Postoperative radiotherapy was given to the patient, during which radiation necrosis occurred. Sixteen months after craniotomy, cranial magnetic resonance imaging revealed recurrence of the tumor. Conclusion Primary intracranial SS is a rare malignant tumor. Primary intracranial SS with hemorrhage and radiation necrosis should be carefully monitored during postoperative radiotherapy. Surgical resection of the tumor combined with postoperative radiotherapy and chemotherapy is currently used, but the prognosis is poor.

Background: We aimed to establish a novel diagnostic model for kidney diseases by combining artificial intelligence with complete mass spectrum information from urinary proteomics. Methods: We enrolled 134 patients (IgA nephropathy, membranous nephropathy, and diabetic kidney disease) and 68 healthy participants as controls, with a total of 610,102 mass spectra from their urinary proteomic profiles. The training data set (80%) was used to create a diagnostic model using XGBoost, random forest (RF), a support vector machine (SVM), and artificial neural networks (ANNs). The diagnostic accuracy was evaluated using a confusion matrix with a test dataset (20%). We also constructed receiver operating-characteristic, Lorenz, and gain curves to evaluate the diagnostic model. Results: Compared with the RF, SVM, and ANNs, the modified XGBoost model, called Kidney Disease Classifier (KDClassifier), showed the best performance. The accuracy of the XGBoost diagnostic model was 96.03%. The area under the curve of the extreme gradient boosting (XGBoost) model was 0.952 (95% confidence interval, 0.9307–0.9733). The Kolmogorov-Smirnov (KS) value of the Lorenz curve was 0.8514. The Lorenz and gain curves showed the strong robustness of the developed model. Conclusion: The KDClassifier achieved high accuracy and robustness and thus provides a potential tool for the classification of kidney diseases. Keywords: Kidney disease classification, urinary proteomics, machine learning algorithm, diagnosis, artificial intelligence

BackgroundAs a neuroprogressive illness, depression is accompanied by brain structural abnormality that extends to many brain regions. However, the progressive structural alteration pattern remains unknown.MethodsTo elaborate the progressive structural alteration of depression according to illness duration, we recruited 195 never-treated first-episode patients with depression and 130 healthy controls (HCs) undergoing T1-weighted MRI scans. Voxel-based morphometry method was adopted to measure gray matter volume (GMV) for each participant. Patients were first divided into three stages according to the length of illness duration, then we explored stage-specific GMV alterations and the causal effect relationship between them using causal structural covariance network (CaSCN) analysis.ResultsOverall, patients with depression presented stage-specific GMV alterations compared with HCs. Regions including the hippocampus, the thalamus and the ventral medial prefrontal cortex (vmPFC) presented GMV alteration at onset of illness. Then as the illness advanced, others regions began to present GMV alterations. These results suggested that GMV alteration originated from the hippocampus, the thalamus and vmPFC then expanded to other brain regions. The results of CaSCN analysis revealed that the hippocampus and the vmPFC corporately exerted causal effect on regions such as nucleus accumbens, the precuneus and the cerebellum. In addition, GMV alteration in the hippocampus was also potentially causally related to that in the dorsolateral frontal gyrus.ConclusionsConsistent with the neuroprogressive hypothesis, our results reveal progressive morphological alteration originating from the vmPFC and the hippocampus and further elucidate possible details about disease progression of depression.