Browse Articles

Network intrusion detection based on improved KNN algorithm

Filling gaps in PM2.5 time series: A broad evaluation from statistical to advanced neural network models

This study addressed the critical challenge of filling gaps in PM2.5 time series data from Pavlodar, Kazakhstan. We developed and evaluated a comprehensive hierarchy of 46 gap-filling methods across five representative gap lengths (5–72 hours), introducing dynamic models capable of adapting to gaps of variable duration. Tree-based models with bidirectional sequence-to-sequence architectures delivered superior performance, with XGB Seq2Seq achieving a mean absolute error of 5.231 ± 0.292 μg/m3 for 12-hour gaps, representing a 63% improvement over basic statistical methods. The advantage of multivariate models incorporating meteorological variables increased substantially with gap length, from modest improvements of 2–3% for 5-hour gaps to significant enhancements of 16–18% for 48–72 hour gaps. Dynamic multivariate models demonstrated remarkable operational flexibility by successfully processing real-world gaps ranging from 1 to 191 hours despite being trained on maximum lengths of 72 hours. Analysis of the reconstructed complete time series revealed that 61.2% of monitored hours exceeded the WHO daily threshold of 15 μg/m3, with strong seasonal patterns and pronounced diurnal cycles. This research advances environmental monitoring capabilities by providing robust methodological tools for addressing data continuity challenges that currently limit the utility of PM2.5 measurements for public health applications and scientific analysis.

Application of a multiple transmitter spacing gradient array TDIP survey in the Huaniushan mining area, Gansu province, China

SpaVGN: A hybrid deep learning framework for high-resolution spatial transcriptomics data reconstruction and spatial domain identification

Spatial transcriptomics has revolutionized the analysis of gene expression while preserving tissue spatial information, which provides novel insights into the cellular composition and function of complex biological tissues. However, current technologies are constrained by limited resolution and data sparsity, compromising the accuracy of downstream analyses. To address these challenges, we developed SpaVGN, a deep learning framework integrating convolutional neural networks, vision transformer, and graph neural networks for high-fidelity gene expression imputation and spatial domain identification. By combining local feature extraction, global attention mechanisms, and spatial graph-based modeling, SpaVGN effectively reconstructs missing transcriptomic data while preserving spatial tissue architecture. Evaluated on melanoma and sagittal posterior mouse brain datasets, SpaVGN outperformed existing methods in gene expression prediction, achieving Pearson correlation coefficients of 0.609 (melanoma) and 0.682 (mouse brain). It clearly delineated tumor regions and lymphoid niches in melanoma tissue, achieving fine-grained resolution of hippocampal subfields, including Cornu Ammonis and Dentate Gyrus, with a Silhouette Score of 0.43 and a Davies-Bouldin Index of 0.86. Validation through UMAP dimensionality reduction and PAGA network analysis demonstrated that SpaVGN significantly mitigates the negative impact of data sparsity in spatial transcriptomics, improving data completeness and spatial continuity. This study presents an innovative solution that enhances the resolution of spatial transcriptomics data, offering cross-tissue applicability and providing a valuable tool for research in biological development, disease, and tumor heterogeneity.

Craniological differentiation amongst Southeast Asian small cats

Nano-polymeric curing agents for enhancing water stability in sandy soils: A sustainable approach for ecological slope protection

The susceptibility of sandy soil slopes to erosion, particularly during rainfall events, poses significant challenges for soil conservation and ecological slope protection. This study explores the potential of nano-polymeric curing agents (NPCA) as a sustainable solution to enhance water stability and slope integrity. Reinforcement depth experiments were conducted to determine the optimal application depth of NPCA, while permeability and erosion tests assessed its impact on water retention and soil stability. Advanced analytical techniques, including scanning electron microscopy (SEM) and Fourier-transform infrared spectroscopy (FTIR), were employed to examine the interactions between NPCA and soil particles. Results show that a 3% NPCA content (mass ratio) achieves the maximum reinforcement depth of 23 mm. Within the optimal reinforcement range (mass ratio < 3%, concentration < 17%), increasing NPCA content enhances soil permeability, reduces the disintegration coefficient, and improves erosion resistance. NPCA encapsulates soil particles, filling pore spaces and binding them through van der Waals forces and hydrogen bonds, forming a durable, elastic membrane that enhances surface stability and water resistance. These findings suggest that NPCA treatment creates a stable, permeable, and breathable environment, crucial for promoting vegetation growth on sandy slopes and offering an effective, sustainable approach to ecological slope protection.

An effective brain stroke diagnosis strategy based on feature extraction and hybrid classifier

Abstract Stroke is a leading cause of death and long-term disability worldwide, and early detection remains a significant clinical challenge. This study proposes an Effective Brain Stroke Diagnosis Strategy (EBDS). The hybrid deep learning framework integrates Vision Transformer (ViT) and VGG16 to enable accurate and interpretable stroke detection from CT images. The model was trained and evaluated using a publicly available dataset from Kaggle, achieving impressive results: a test accuracy of 99.6%, a precision of 1.00 for normal cases and 0.98 for stroke cases, a recall of 0.99 for normal cases and 1.00 for stroke cases, and an overall F1-score of 0.99. These results demonstrate the robustness and reliability of the EBDS model, which outperforms several recent state-of-the-art methods. To enhance clinical trust, the model incorporates explainability techniques, such as Grad-CAM and LIME, which provide visual insights into its decision-making process. The EBDS framework is designed for real-time application in emergency settings, offering both high diagnostic performance and interpretability. This work addresses a critical research gap in early brain stroke diagnosis and contributes a scalable, explainable, and clinically relevant solution for medical imaging diagnostics.

A scoping review of the levels, implementation strategies, enablers, and barriers to cervical, breast, and colorectal cancer screening among migrant populations in selected English-speaking high-income countries

Background Cancer remains one of the leading causes of mortality and morbidity worldwide with colorectal, cervical, and breast cancers accounting for significant proportion of preventable deaths. Early screening, diagnosis, and treatment could prevent many of these deaths. However, migrants face persistent disparities in the screening, early diagnosis, and treatment of these cancers. This study synthesizes evidence on cancer screening uptake, implementation strategies, as well as their enablers and barriers among migrants in English-speaking high-income countries (Australia, the USA, the UK, Canada, and New Zealand). Methods We conducted a scoping review of studies published in any language between 1 January 2015 and 31 December 2024. Studies were retrieved from four databases: PubMed, Scopus, Embase, and Web of Science. Search terms were developed based on four domains: types of cancer (colorectal, cervical, and breast), migrant populations, screening coverage, and country of residence. The uptake of cancer screening among migrants in selected countries was determined. A thematic analysis was conducted to analyze the data and identify key themes related to the implementation of cancer screening strategies, as well as their enablers and barriers. Results A total of 80 studies were included in the review. Migrants exhibited varied levels of utilization of cancer screening such as cervical cancer (41% − 84%), breast cancer (24%−87%), and colorectal cancer (4%−55%). Four themes related to the implementation of cancer screening strategies were identified: i) culturally tailored health education and communication, ii) trust-building initiatives with providers and health systems, iii) family and community support for acculturation and engagement, iv) awareness and knowledge on increased risk perception. Several barriers to the implementation of cancer screening strategies were identified, including lack of insurance, transportation challenges, difficulty in speaking and understanding English, inflexible work hours of health services, cultural taboos, stigma, poverty, and undocumented (illegal) status of migrants. Enablers of the implementation of cancer screening strategies included faith-based messaging on cancer screening, community partnerships, home-based fecal immunochemical test kits, availability of after-hours services, gender-concordant care, social networks, acculturation, and trust-building. Conclusions The uptake of cancer screening (breast, cervical, colorectal) varied and had low among migrants (e.g., refugees, culturally and linguistically diverse populations). Targeted, culturally tailored approaches, expanding interpreter services, and fostering cross-sector collaborations (e.g., linking screenings to cultural events) are essential for addressing disparities in cancer screening among migrants. Culturally sensitive and adaptive, equity-focussed interventions on cancer screening should be prioritized by ensuring sustained funding, disaggregated data collection on the uptake of cancers screening and design and implementation of programs on targeting diverse population groups.

Levitating platform could ride sunlight into the ‘ignorosphere’

Uniaxial compressive strength of cemented coal gangue backfill based on response surface methodology

Seroprevalence of antibodies against hepatitis A and E among the general population in 5 provinces, Lao People’s Democratic Republic: Variation according to location

Hepatitis A and E viruses (HAV and HEV) are transmitted through the faecal-oral route: via contaminated food, water, and contact with infected people and/or animals for HEV. Due to limited data from Lao People’s Democratic Republic (Lao PDR), we assessed HAV and HEV seroprevalence in the Lao general population. A cross-sectional study collected 2412 serum samples and demographic information from participants (5–93 years) across five provinces. Anti-HAV (IgM and IgG) and anti-HEV antibodies (IgG) were detected by enzyme-linked immunosorbent assay (Dia.Pro). The overall seroprevalence of anti-HAV was 84.3% and anti-HEV was 57.9%. Seropositivity was associated with occupation, location, increasing age, ethnicity (only for anti-HAV) and sex (only for anti-HEV). The age at which 50% of the population was seropositive differed from 12 years (Oudomxay) to 26 years (Savannakhet and Vientiane) for anti-HAV and from 22 years (Savannakhet) to 49 years (Vientiane) for anti-HEV. The prevalence of double seropositivity was high overall (53.4%), particularly in Savannakhet and Champasack. These significant differences according to location and socio-demographics may be the result of variation of exposure to the viruses, such as through water, sanitation and hygiene-related risks, occupational exposure and animal contact. Further studies are warranted to identify the most important risks for transmission in Lao PDR in order to develop targeted public health interventions.

Spatial correlation in economic analysis of climate change

Structural variation in 1,019 diverse humans based on long-read sequencing

Abstract Genomic structural variants (SVs) contribute substantially to genetic diversity and human diseases 1–4 , yet remain under-characterized in population-scale cohorts 5 . Here we conducted long-read sequencing 6 in 1,019 humans to construct an intermediate-coverage resource covering 26 populations from the 1000 Genomes Project. Integrating linear and graph genome-based analyses, we uncover over 100,000 sequence-resolved biallelic SVs and we genotype 300,000 multiallelic variable number of tandem repeats 7 , advancing SV characterization over short-read-based population-scale surveys 3,4 . We characterize deletions, duplications, insertions and inversions in distinct populations. Long interspersed nuclear element-1 (L1) and SINE-VNTR-Alu (SVA) retrotransposition activities mediate the transduction 8,9 of unique sequence stretches in 5′ or 3′, depending on source mobile element class and locus. SV breakpoint analyses point to a spectrum of homology-mediated processes contributing to SV formation and recurrent deletion events. Our open-access resource underscores the value of long-read sequencing in advancing SV characterization and enables guiding variant prioritization in patient genomes.

Effects on muscular activity and usability of soft active versus rigid passive back exoskeleton during symmetric lifting tasks

FedNolowe: A normalized loss-based weighted aggregation strategy for robust federated learning in heterogeneous environments

Federated Learning supports collaborative model training across distributed clients while keeping sensitive data decentralized. Still, non-independent and identically distributed data pose challenges like unstable convergence and client drift. We propose Federated Normalized Loss-based Weighted Aggregation (FedNolowe) (Code is available at https://github.com/dongld-2020/fednolowe), a new method that weights client contributions using normalized training losses, favoring those with lower losses to improve global model stability. Unlike prior methods tied to dataset sizes or resource-heavy techniques, FedNolowe employs a two-stage L1 normalization, reducing computational complexity by 40% in floating-point operations while matching state-of-the-art performance. A detailed sensitivity analysis shows our two-stage weighting maintains stability in heterogeneous settings by mitigating extreme loss impacts while remaining effective in independent and identically distributed scenarios.

Dynamic removal of methylene blue and methyl orange from water using biochar derived from kitchen waste

Intergenerational instructional strategies and elderly preferences for digital applications: A Malaysian case study

Introduction As Malaysia transitions into an ageing society, older adults increasingly face challenges in acquiring digital literacy, which impacts their ability to engage with essential online services, financial transactions, healthcare applications and communication platforms. The percentage of individuals aged 65 and above in Malaysia rose from 7.2% in 2022 to 7.4% in 2023, highlighting the growing needs for digital inclusion among this demographic. While internet adoption among older individuals is increasing, many still struggle due to psychological, cognitive, and physical barriers. Factors such as low self-efficacy, fear of complexity, and age-related physical limitations hinder their effective use of digital applications. Despite government initiatives such as MyDigital and the Malaysia Digital Economy Blueprint, gaps in digital literacy persist among older adults. Research suggests that intergenerational programs (IPs), where younger individuals assist older adults in learning digital skills, can bridge this gap. These programs promote collaborative learning, reduce social isolation, and foster meaningful intergenerational relationships. However, existing instructional strategies within IPs often fail to accommodate the specific learning needs and preferences of older adults, limiting their effectiveness. Addressing these instructional gaps is essential to ensuring older adults’ successful integration into the digital world. Aims This study aims to investigate the preferences of older adult individuals in Malaysia regarding digital applications and to propose effective intergenerational instructional strategies that enhance their digital learning experiences. Methods This qualitative study explores older adults’ preferences for digital applications and their experiences with intergenerational instructional strategies in Malaysia. A total of 26 older adults and 13 young instructors participated in digital applications workshops at two Pusat Aktiviti Warga Emas (PAWE) centres on August, 2024. Older adults were paired with young instructors in small groups to guide them in using digital applications through hands-on learning. Data collection involved semi-structured interviews before and after the sessions, as well as observations. Thematic analysis is used to analyse interview data, identifying key insights into improving digital learning for older adults in Malaysia. Results The study, conducted at two PAWE centres in Perak, Malaysia, explored intergenerational digital learning between 26 older adults (ages 59–82) and 13 young instructors (ages 20–24). Findings highlight the importance of structured instructional strategies, including interactive learning, direct instruction, and clear communication, to enhance digital literacy among older adults. Personalized learning approaches, small group discussions, and adapting content to individual needs improve engagement and comprehension. Smartphone familiarity plays a key role, with participants favouring WhatsApp and Facebook for communication. Older adults face challenges such as fear of complexity, physical limitations, and security concerns, while young instructors benefit from training to improve communication and instructional skills. The program fosters confidence, intergenerational bonding, and digital inclusion, demonstrating the value of tailored learning strategies in bridging the digital divide.

Optimization of the microstructural, mechanical, and radiation shielding properties of Al-30B4C-25 W hybrid composites with Gd2O3 reinforcement

Infertility screening in unmarried women: A scoping review protocol

Objective This scoping review aims to systematically map the landscape of infertility screening in unmarried women. Introduction Infertility screening in unmarried women represents a significant and often neglected area within reproductive health. This population faces unique challenges and barriers, including social stigma, cultural norms, and limited access to care, making a comprehensive understanding of current screening practices essential. Inclusion criteria Study selection will be guided by the PCCT framework (Population, Concept, Context, and study type), considering diverse study designs (quantitative, qualitative, mixed-methods, reviews) and grey literature, focusing on infertility screening in unmarried women. Methods This scoping review will follow the PRISMA-ScR guidelines and utilize a 14-step framework based on Arksey and O’Malley’s methodology, incorporating enhancements by Tricco and Peters. A comprehensive search strategy will be employed, using controlled vocabulary and free-text method. Databases of Web of Science (ISI), PubMed, Scopus and search engines like Google Scholar will be searched, and supplemented by forward and backward citation tracking. Inclusion/exclusion criteria will be applied iteratively. Two independent reviewers will screen titles/abstracts and full texts, resolving disagreements through consensus or a third reviewer. Data will be charted using a predefined template, and findings will be presented in tables and diagrams, accompanied by a narrative synthesis. Discussion This scoping review will provide a comprehensive overview of the current state of knowledge regarding infertility screening in unmarried women, a significantly under-researched area. The findings will be critical for informing the development of culturally sensitive guidelines, targeted interventions, and future research to address the reproductive health needs of this underserved population.

Research on cause analysis and management of coal mine safety risk based on social network and bow-tie model

Abstract Accurate identification of coal mine safety risks is a crucial foundation for mitigating coal mine disasters. This study integrates social network analysis (SNA), the bow-tie model, and association rule mining to systematically analyze safety accident data from a coal mine. A total of 85 causative factors were extracted from 72 accidents and assessed through frequency, marginal influence, and centrality indicators to identify key risk contributors. The bow-tie model was employed to structure these causes into a safety risk control framework based on preventive and mitigation measures. Furthermore, the Apriori algorithm was applied to uncover hidden associations among gas safety risk factors, revealing critical compound relationships among factors such as inadequate safety management, insufficient inspections, high incidence of “three violations”, and poor safety education. The findings indicate that management and human-related factors, particularly the absence of effective safety management systems, safety violations, and inadequate training, are the primary contributors to accidents in coal mines. Consequently, it is imperative to address these issues collectively to ensure effective risk prevention in such environments. The coal mine safety risk causality control model established in conjunction with the butterfly diagram model holds significant theoretical and practical value for coal mine safety production.