Browse Articles

Development and validation of a clinical scoring system for predicting endoscopic failure in upper gastrointestinal bleeding ulcer: a case control study

MS-EGT-Net: a multi-scale enhanced graph-transformer network for diabetic foot ulcer classification

Multifunctional stanene driven by a magnetic topological switch

Machine learning-based prediction of overall survival in non-metastatic renal cell carcinoma after radical nephrectomy

Physicochemical properties and DFT insight into choline chloride-levulinic acid deep eutectic solvent and its binary mixtures with carboxylic acids

Estimating carbon prices for revenue parity between timber and carbon in forest plantations

Landslide detection in remote sensing images on the basis of feature recursion and sample balance

Discovery and quantification of novel diazepam metabolites BP-246 and BP-271 in freshwater fish using HRMS and UPLC-MS/MS

Detection and antimicrobial susceptibility of genital mycoplasmas in women and men, with 23S rRNA and parC mutation analysis in Mycoplasma genitalium

A lightweight anonymous authentication scheme for federated learning

Abstract Federated learning enables collaborative model training between central servers and distributed clients without collecting users’ raw sensitive data, which effectively promotes the large-scale deployment of intelligent collaborative services. Considering the high sensitivity of local training data and model gradient parameters in federated learning, protecting identity privacy and interaction security has become extremely critical. Therefore, mutual identity authentication is indispensable to restrict illegal client access and prevent malicious parameter transmission and data tampering. In this paper, we propose a lightweight anonymous authentication scheme for federated learning (FedLAS), which realizes secure mutual authentication between servers and clients and establishes a shared session key for subsequent encrypted interaction. In particular, the proposed scheme eliminates the reliance on high-cost cryptographic operations such as bilinear pairing, thus minimizing computational and communication overhead. Furthermore, informal security analysis demonstrates that our FedLAS scheme can resist multiple common attacks and meet predefined security requirements. Extensive comparative experiments show that the FedLAS scheme achieves excellent performance in computational and communication cost. It is well suitable for resource-constrained federated learning scenarios.

A unified closed-loop stabilization framework for fixed-wing UAVs via natural-selection-enhanced multi-objective particle swarm optimization

Behavioural responses of Neotropical raptors to drone approaches to nests

Modeling grain selection for single-crystal thin-wall inconel 718 tubes by laser powder bed fusion

Fabrication of low-cost ceramic microfiltration membranes with the addition of fly ash to kaolin-natural zeolite structure for greywater treatment

Identification of a sequence element regulating H3K9 methylation at the ap2-g locus in Plasmodium falciparum

Hybrid multilayer perceptron models optimized by evolutionary algorithms for urban air quality forecasting: a case study of Shiraz, Iran

Multi-scale petrographic and geotechnical characterization of Kuldana formation limestone and implications for asphalt concrete stiffness and fatigue design

Abstract This study evaluates the suitability of Kuldana Formation limestone from Nammal Gorge, Salt Range–Potwar Basin, Pakistan, for construction and asphalt pavement applications through integrated petrographic, geotechnical, durability, statistical, and pavement-performance analyses. Fresh, unweathered limestone blocks were collected from intact outcrops and examined using optical microscopy and SEM–EDS to identify mineralogical composition, calcite veining, pore structure, vuggy porosity, and microfracture characteristics. Standardized EN/ISRM/ASTM-aligned laboratory tests were performed to determine mechanical properties, including unconfined compressive strength, unconfined tensile strength, point load index, shear strength, and Schmidt rebound, as well as physical and durability indicators such as specific gravity, bulk density, porosity, water absorption, crushing strength, Los Angeles abrasion coefficient, and freeze–thaw resistance. The limestone showed considerable variability, with UCS ranging from 25 to 98 MPa, UTS from 20 to 80 MPa, water absorption from 0.30 to 0.72%, specific gravity from 1.67 to 2.88, porosity from 1.13 to 2.78%, LA coefficient from 10.31 to 35.20, and freeze–thaw resistance from 14.2 to 60.5. This variability is mainly related to calcite veining, pore–fracture connectivity, vuggy porosity, and microcrack density. Pearson correlation and regression analyses identified coherent relationships among selected physical, mechanical, and durability parameters, while scattered data points reflected defect-controlled responses. Pavement-performance assessment of limestone-aggregate high-stiffness asphalt concrete mixtures showed that the mixtures can satisfy KR7 fatigue requirements and may support limited asphalt-layer thickness optimization, subject to field validation and production quality control. Overall, Kuldana Formation limestone shows promising potential as a construction and asphalt pavement aggregate.

Enhanced MRI brain tumor segmentation with DNet and hybrid dice-weighted cross-entropy loss

Abstract Localization of brain tumors via magnetic resonance imaging (MRI) is critically important for diagnosis, treatment, and surgical operations. This study proposes DNet, an optimized and highly efficient deep learning framework for the automatic detection and segmentation of brain tumors from MR images. On the basis of the classical UNet structure, DNet reduces the risk of overfitting by balancing the integration of batch normalization and dropout layers; it also achieves higher accuracy in small tumor regions by combining Dice loss and weighted cross-entropy (WCE) loss. To improve the model’s robustness and predictive stability, an ensemble and test-time augmentation (TTA) were applied during testing. The dataset used in the study consisted of 1664 MR images containing three tumor classes: glioma, meningioma, and pituitary. Data preprocessing and augmentation steps, including contrast correction, reflection, and cropping, were applied to the images. The model was trained via the five-fold cross-validation method; the mean intersection-over-union (mIoU) and accuracy were used for evaluation. The experimental results demonstrate improved performance over the classical UNet. The DNet model achieved competitive performance compared with existing approaches in the literature, reaching 99.6% accuracy and a test mIoU value of 0.8617 on a three-class dataset. Furthermore, ablation studies revealed that reducing the number of classes increased the model’s mIoU by approximately 3%. In conclusion, the proposed DNet framework provides a robust and efficient framework for medical image segmentation. In addition to the primary dataset, consistent performance was also observed on an independent dataset, supporting the robustness of the proposed approach. Nevertheless, further validation on diverse and multi-center datasets is required for real-world clinical deployment.

Research on the dry screening characteristics of fine coal based on the discrete element method

Phased pruning in neural networks recapitulates selectivity–fragility trade-offs in brain development