Browse Articles
Discover research articles across all indexed journals
A deep learning approach to predict temporal changes of subdural hemorrhage on computed tomography
Abstract Subdural hemorrhage (SDH) is a critical condition requiring prompt assessment of its progression using computed tomography (CT). This study aimed to develop a deep-learning model to predict temporal changes in SDH by leveraging Hounsfield Units (HU) to estimate hemorrhage age across acute, subacute, and chronic stages. A total of 825 pre-processed CT slices from the RSNA dataset were balanced across SDH stages and analyzed using a convolutional neural network (CNN) implemented in Python on Google Colab. Model performance was evaluated using accuracy, sensitivity, specificity, precision, F1-score, and the Area Under the Receiver Operating Characteristic Curve (AUC-ROC). The model achieved 83.11% training accuracy and 85.33% prediction accuracy. Sensitivity for acute, subacute, and chronic SDH was 86.67%, 84%, and 85.33%, respectively, with specificity values of 94%, 88%, and 96%. Precision scores were 87.84%, 77.78%, and 91.43%, while F1 scores were 87.25%, 80.77%, and 88.28%. AUC-ROC values ranged from 0.9394 to 0.9731 across five folds, reflecting robust classification performance. The results highlight the model’s potential to support radiologists as a second-reader tool, streamline emergency triage, and enhance diagnostic efficiency within clinical workflows.
Real-time prediction of soil bearing capacity in clayey soils using drilling parameters and statistical modeling
Stiffness optimization of electric spindle performance based on multi-layer perceptron integrated Bayesian
Characterising protective immune responses to SARS-CoV-2 in urban and rural Malawi between February 2021 and April 2022
Abstract SARS-CoV-2 transmission in Malawi remains unclear due to high proportions of mild/asymptomatic infections and limited diagnostics. Existing seroprevalence studies in Malawi have primarily used convenience samples and enzyme-linked immunosorbent assays (ELISAs). We assessed SARS-CoV-2 neutralisation in a longitudinal Malawian population-based cohort, assessing protective immunity post-infection and vaccination. Sera were obtained from rural (Karonga, n = 958) and urban (Lilongwe, n = 918) based participants at three-monthly intervals (February 2021-April 2022). Neutralising antibodies against SARS-CoV-2 were measured using human immunodeficiency (HIV)-based pseudotype assays in HIV-uninfected participants, and vesicular stomatitis virus-based assays in HIV-infected participants and an HIV-uninfected subset. Nucleocapsid ELISAs identified vaccinated participants also infected. SARS-CoV-2 neutralisation profiles increased in complexity over time from rising vaccination coverage and emerging variants. Neutralising antibody prevalence was higher in Lilongwe than Karonga (68.1% (CI 63.5–72.4) vs. 45.4% (CI 41.6–49.3), Survey 4). Hybrid immune and solely vaccinated participants exhibited higher titres than those solely infected. Children < 15 years had the lowest neutralising antibody titres among infected (not vaccinated) participants. People living with HIV had lower neutralising responses than those HIV-uninfected, particularly post vaccination. We therefore recommend surveillance of children and people living with HIV as low neutralisation responses increase reinfection risk. COVID-19 vaccination should be prioritised for HIV-infected individuals.
Interplay between key metabolic hormones, metabolic factors, renal function, and heart rate variability in humans with obesity
Abstract This study aimed to provide the first integrative assessment of clinical, metabolic, renal, hormonal, and heart rate variability (HRV) parameters in individuals with obesity, stratified by metabolic syndrome (MetS) and insulin resistance (IR), clarifying shared and distinct mechanisms beyond prior HRV- or hormone-focused studies. Among 45 participants with obesity (BMI ≥ 25 kg/m²), 67% had IR and 42% had MetS. Both groups exhibited increased %fat, triglyceride, leptin, and resting heart rate, with decreased QUICKI and high-density lipoprotein cholesterol (HDL-C) ( p < 0.05 all). Frequency-domain HRV (VLF and LF ms²) and overall variability (SDNN) were significantly decreased in the IR group ( p < 0.05 all). Leptin showed significant positive correlations with obesity, IR, and creatinine clearance ( p < 0.05 all). Adiponectin exhibited positive correlations with hip circumference, HDL-C, and pNN50 while HDL-C showed negative correlations with the number of MetS criteria, obesity, IR, and leptin, but positive correlations with parasympathetic HRV (SDSD and RMSSD) ( p < 0.05 all), suggesting a protective role across multiple systems. Creatinine clearance and eGFR revealed positive correlations with parasympathetic HRV (HF nu) and negative correlations with sympathetic HRV (LF nu and LF/HF ratio). In conclusion, this study underscores the complex interplay between these systems, enhancing our understanding of their shared and distinct mechanisms.
A prognostic risk prediction model for gastric cancer based on the EFNA4 and ETS1 regulatory axis in tumor cells
Abstract Gastric cancer (GC) is a major cause of cancer-related deaths worldwide, and is characterised by intricate molecular mechanisms. However, analysis of its molecular and clinical characteristics is complicated by its histological and etiological heterogeneity. Dysregulation of the PI3K-Akt signalling pathway is common in GC. In this study, we have identified the hub gene Ephrin A4 ( EFNA4 ) in the PI3K-Akt pathway based on transcriptome data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases, and ETS Proto-Oncogene 1 ( ETS1 ) was a gene related to EFNA4. Using publicly accessible datasets, we conducted bioinformatics analyses to evaluate the expression profiles, functional roles, and prognostic significance of EFNA4 and ETS1 and further explored their relationship in GC. Subsequently, consensus clustering was performed on 373 TCGA-STAD datasets based on the expression matrices of EFNA4 and ETS1 to assess their interconnections with relevant signalling cascades and immune system components. To address the challenges posed by tumour heterogeneity and reveal the expression patterns of EFNA4 and ETS1 in GC tissues, we performed reanalysis of single-cell RNA sequencing (scRNA-seq) data of GC samples. We constructed a tumour-based risk signature for GC based on EFNA4, ETS1, and marker genes of the tumour cell cluster. The prognostic value of the prognosis prediction model was verified using TCGA database to facilitate the clinical application of tumour cell features in GC prognosis. Our study reveals EFNA4 and ETS1 expression patterns in GC, implicating their roles in pathogenesis. An integrated EFNA4-ETS1 prognostic model improves GC risk stratification. Although EFNA4 has been shown to promote metastasis in liver cancer, the contradictory mechanism of its high expression and good prognosis in GC remains to be elucidated, which may involve its antagonistic effects with ETS1, and requires further exploration.
Metagenomic next-generation sequencing reveals respiratory flora distribution in COVID-19
Rasch analysis and differential item functioning of the marital happiness scale in Oman
Breast cancer awareness and its predictors among university students and employees in Asir region, Saudi Arabia: a cross-sectional study
The impact of precipitation on ANC service utilisation and healthcare access in Malawi
Abstract Malawi is vulnerable to climate-related shocks, which are projected to worsen. Whilst some dimensions of this vulnerability have been characterised, little is known about healthcare sector resilience. Coupling facility-specific data on antenatal care (ANC) service provision in Malawi with gridded precipitation data from 2012-2024 we use linear regression analyses to characterise the historic relationship between precipitation and healthcare access. We estimate that precipitation negatively impacted ANC service utilisation in Malawi, with up to 1 in 20 appointments disrupted annually in some districts. Projecting further to 2060 indicates that, cumulatively, up to 250,000 pregnancies could be affected. Notably, if precipitation patterns from 1941 to 1953 had persisted into the 21st century, disruptions between 2012 and 2024 would be a hundred times less frequent, highlighting the significant influence of anthropogenic climate change on healthcare access. In a country already facing high maternal and neonatal mortality, such disruptions could further hinder access to care and worsen health outcomes. To mitigate this, interventions should focus on preserving or improving the physical accessibility of facilities, particularly through resilient transport services and road networks.
Filtering out mislabeled training instances using black-box optimization and quantum annealing
Abstract This study proposes an approach for removing mislabeled instances from contaminated training datasets by combining surrogate model-based black-box optimization (BBO) with postprocessing and quantum annealing. Mislabeled training instances, a common issue in real-world datasets, often degrade model generalization, necessitating robust and efficient noise-removal strategies. The proposed method evaluates filtered training subsets based on validation loss, iteratively refines loss estimates through surrogate model-based BBO with postprocessing, and leverages quantum annealing to efficiently sample diverse training subsets with low validation error. Experiments on a noisy majority bit task demonstrate the method’s ability to prioritize the removal of high-risk mislabeled instances. Integrating D-Wave’s clique sampler running on a physical quantum annealer achieves faster optimization and higher-quality training subsets compared to OpenJij’s simulated quantum annealing sampler or Neal’s simulated annealing sampler, offering a scalable framework for enhancing dataset quality. This work highlights the effectiveness of the proposed method for supervised learning tasks, with future directions including its application to unsupervised learning, real-world datasets, and large-scale implementations.
Benchmarking diffusion models against state-of-the-art architectures for OCT fluid biomarker segmentation
Objectives Retinal diseases, major causes of vision impairment and blindness, are assessed using optical coherence tomography (OCT) scans. Automated report generation for retinal OCT scans, powered by deep learning, can help standardize interpretations and track retinal disease in clinical settings. A key challenge is accurately segmenting retinal disease signatures. This study explores using the diffusion model to segment subretinal fluid (SRF), intraretinal fluid (IRF), and pigment epithelial detachment (PED) in typical clinical settings, comparing their performance to other leading segmentation models. Methods We labeled OCT scans and extracted those with specific pathologic retinal features: 269 scans with SRF, 224 scans with IRF, and 114 scans with PED. Three trained reviewers manually segmented these features for downstream analysis. Using manually segmented scans as the ground truth, we trained the diffusion model, Nested U-Net, nnU-Net, TransUNet, and SwinUNet to predict these segmentations. All models were evaluated using 5-fold cross-validation, with performance measured by Dice coefficient, sensitivity, specificity, Pearson correlation coefficient, and R 2 . Results All models show high similarly with ground truth segmentations in predicting SRF, IRF, and PED, as shown by the Dice coefficient (Diffusion model: 0.81 ± 0.12, 0.66 ± 0.09, 0.75 ± 0.11). The diffusion model has relatively higher sensitivity compared to most other models, while all models display very high specificity. The Pearson correlation coefficient and R 2 values show strongly associated pixel quantification of segmented areas for models, with the nnU-Net model performing the strongest overall. Conclusion This study demonstrates that while diffusion models can comparably segment retinal pathologies using a limited number of manually annotated scans, the nnU-Net model remains the most effective overall for automated OCT analysis.
Mechanistic analysis of luteolin in mitigating dry age-related macular degeneration through network pharmacology and experimental validation
‘You decided I am old enough for the transition, but not old enough to have a say?’: Exploring young people’s, parents’, and healthcare providers’ views and experiences of Type 1 Diabetes paediatric to adult healthcare transition in Saudi Arabia
Background Young people with Type 1 diabetes in Saudi Arabia transition from paediatric to adult care at a culturally defined age of 14, which is younger than the average transition age in Western societies. The aim of this study was to elicit the experiences of young people with Type 1 diabetes, their parents, and healthcare providers in Saudi Arabia as they transitioned from paediatric to adult care. Methods In healthcare, Design Thinking is a human-centred approach that draws on participants’ experiences and perspectives to design and develop interventions, models, or services that meet the needs of stakeholders. This study reports the first inspiration phase of the Design Thinking process. Four parallel exploration workshops were held with pre- and post-transition young people with Type 1 diabetes (n = 12), their parents (n = 8), and healthcare providers (n = 7). Findings Six key themes were identified from the workshops’ data analysis. For young people, the key themes were facing the unknown and preparedness; developing autonomy and recognition as an independent person; and interacting with the adult healthcare team. For parents, the themes were navigating the shift in parental role and involvement in care, interacting with healthcare professionals, and changing support needs. For healthcare providers, the key theme was balancing independence and care approaches. Conclusion The Inspiration phase of the Design Thinking approach provided valuable insights from the healthcare transition experiences of young people with Type 1 diabetes, their parents, and healthcare providers in Saudi Arabia. The generated insights facilitated the identification of areas for interventions in the process’s following phases.
A robust framework for evaluating green mines towards sustainable development
Abstract The development of green mines is essential for promoting sustainability in the mining sector due to the significant ecological impacts of resource extraction. This study proposes a novel hybrid multi-criteria decision-making (MCDM) framework that integrates Spherical Fuzzy Sets (SFSs) with SWOT analysis, the CRITIC method, and Grey Relational Analysis (GRA). The framework introduces several innovations: it applies SFS-based MCDM for the first time to green mine evaluation in Egypt, structures 37 sustainability-related criteria under SWOT dimensions, and employs SF-CRITIC for objective weighting without subjective comparisons. The model is applied to assess 20 gold mines, where the SF-GRA method is used to rank alternatives based on proximity to an ideal solution. The results show that GME20 consistently ranks highest, while GME5 ranks lowest. A sensitivity analysis is conducted by varying the Grey relational coefficient and simulating 37 weight scenarios, demonstrating stable rankings and strong model resilience. Comparative analysis against ten SFS-based MCDM methods confirms the consistency of results, with Spearman correlation coefficients exceeding 0.77. In addition to its methodological novelty, the framework supports interpretable decision outcomes by identifying key sustainability drivers such as renewable energy adoption and land reclamation. This contributes actionable insights for policymakers and stakeholders, enabling informed green investment and regulatory decisions. The study offers a transparent, reproducible, and scalable tool for sustainability evaluation in resource-intensive industries. The proposed model introduces a structured integration of SWOT-based criteria classification, objective weight computation via SF-CRITIC, and robust alternative ranking using SF-GRA. Furthermore, it contributes uniquely by applying the methodology to the underexplored context of green mine evaluation in Egypt. These distinctions articulate the methodological and application-based novelties of the proposed framework.
Sociodemographic and cultural factors are related to singlehood rates: A multilevel analysis across 59 countries from the World Values Survey
Singlehood, which refers to remaining without a lifetime partner, has become an increasingly common phenomenon. However, there is still limited understanding of the individual-level sociodemographic and country-level cultural factors that predict one’s singlehood status. We addressed this question by utilizing data from the World Values Survey, which included responses from 71,169 individuals across 59 countries. Through multilevel modeling, we discovered that several factors increase the likelihood of being single. These factors include being younger, being male, residing in a larger town, having a higher level of education, having a lower income, being unemployed, and living in countries characterized by higher individualism and lower flexibility. Additionally, the likelihood of being single varied according to country-level individualism and flexibility, interacting with various individual-level factors. These findings suggest that the significance of individual sociodemographic characteristics on the prevalence of single individuals depends on country-level traits related to individualism-collectivism and flexibility-monumentalism.
Analysis of the matching of community vitality and service facility supply and demand in the pedestrian bridge areas of Beijing
The Covid-19 pandemic in Sweden: Prolonged and unevenly distributed effects on the volume of pediatric anesthesia and surgery demonstrated by data from the Swedish Perioperative Register
Background In 2020, Covid-19 pushed Swedish health care to its limits regarding access to hospital beds and staffing. A previous investigation of the effects of the first wave of the pandemic in the spring of 2020 revealed a substantial reduction in elective pediatric surgery. The aim of the present study was to expand this analysis on a national and regional level during almost three years with Covid-19. Methods For this retrospective cohort study, routine data from all procedures in patients <16 years of age in 2019–2022 were extracted from the Swedish Perioperative Register. Data were analyzed according to level of care, type of surgery, procedure code and emergency or elective surgery. Results During 2020–2022, the number of surgeries registered was 19,944 fewer than expected as compared to pre-pandemic levels, i.e., a reduction of about 12%. Elective surgery showed a total reduction of 17% while emergency surgery was unaffected. The most dramatic decrease was found in county hospitals where elective surgery was reduced by 28% and the largest effect was found in Ear, Nose, and Throat/oral surgery (−34%). Patient age at the time of surgery did not show any notable differences in total, except for grommets insertion in 2021 and adenoidectomy in 2021 and 2022 compared to 2019. Conclusion The Covid-19 pandemic affected the number of surgical procedures in children for more than two years. Future studies of the long-term effects of the large number of canceled operations are warranted.