Search Results (3,644)

Cough is a key symptom reflecting respiratory health, with its frequency and pattern providing valuable insights into disease progression and clinical management. Objective and reliable cough detection systems are therefore of broad significance for healthcare and remote monitoring. However, existing algorithms often struggle to jointly model spatial and temporal information, limiting their robustness in real-world applications. To address this issue, we propose a cough recognition framework based on frequency-modulated continuous-wave (FMCW) radar, integrating a deep convolutional neural network (CNN) with a Self-Attention mechanism. The CNN extracts spatial features from range-Doppler maps, while Self-Attention captures temporal dependencies, and effective data augmentation strategies enhance generalization by simulating position variations and masking local dependencies. To rigorously evaluate practicality, we collected a large-scale radar dataset covering diverse positions, orientations, and activities. Experimental results demonstrate that, under subject-independent five-fold cross-validation, the proposed model achieved a mean F1-score of $0.974\pm 0.016$ and an accuracy of $99.05\pm 0.55$ %, further supported by high precision of $98.77\pm 1.05$ %, recall of $96.07\pm 2.16$ %, and specificity of $99.73\pm 0.23$ %. These results confirm that our method is not only robust in realistic scenarios but also provides a practical pathway toward continuous, non-invasive, and privacy-preserving respiratory health monitoring in both clinical and telehealth applications. Full article

To reduce the renewable energy waste and carbon emissions predicted for the current expansion plan, this study proposes a hierarchical collaborative optimization model for the planning of generation and transmission expansion plan in cross-regional power systems considering energy storage and load transfer. In the upper layer, the upper limit of expansion is determined according to China’s current policy and expansion plan for the power system. This level completes the annual power expansion plan and provides scale data of power generation facilities and supporting infrastructures for the lower level. The lower layer is the operation level, which simulates the operation of the power system throughout the year. To find the defects of the current plan and provide an optimization scheme, the optimization model is used to analyze China’s power system in 2030. The utilization of renewable energy and power facilities is analyzed, along with the carbon emissions. An improved power expansion plan that comprehensively considers energy storage, transmission and load transfer for China’s carbon peak is proposed. The proposed scheme increases the utilization rate of renewable energy to 97.058%, reduces CO₂ emissions by 224 million tons, and reduces the installed capacity of thermal power by about 18.686 million kilowatts, verifying the effectiveness of the scheme. Full article

Rapid urbanization in China has intensified spatial and social disparities between urban and rural areas, posing major challenges to sustainable rural development. Traditional top-down rural construction and evaluation models often neglect villagers’ everyday practices, resulting in mismatches between spatial planning and actual use. This study develops a cross-scale, bottom-up framework for assessing rural construction through social network analysis (SNA), taking Xiongfan Village in Dawu County, Hubei Province, as a case study. At the village scale, the comparison between the “Public Space Structure Network” and the “Villagers’ Space Usage Behavior Network” reveals a significant mismatch between spatial compactness and behavioral dispersion, with high-frequency activities concentrated along the north–south axis while peripheral and east–west spaces remain underutilized. At the township scale, GPS-based analysis shows that the revitalization of Xiongfan transformed it from a peripheral node into a central hub, restructuring the network into a new pattern of “characteristic towns—traditional villages—ecological scenic areas.” These findings highlight the dual role of rural construction in both meeting residents’ daily needs and fostering regional integration. The proposed cross-scale SNA framework not only advances methodological tools for evaluating rural construction but also provides practical guidance for inclusive, resilient, and sustainable urban–rural development in line with the UN Sustainable Development Goals (SDGs). Full article

Background/Objectives: Fibromyalgia (FM) is a chronic pain syndrome frequently associated with severe pain, sleep disturbances, worry, and depressive rumination. Although previous studies have shown links among these factors, no study has specifically examined the mediating role of sleep disturbances in the relationship between forms of Repetitive Negative Thinking (i.e., worry and rumination) and pain intensity. This study aimed to investigate whether sleep disturbances mediate the relationship between: (1) worry and pain intensity and (2) depressive rumination and pain intensity. Methods: An online cross-sectional survey was conducted with a sample of 867 Italian adults who reported having received an FM diagnosis from a rheumatologist or pain physician. After screening, 733 participants (97.3% female; mean age = 51.0 ± 9.95 years) were included in the analyses. Participants completed the Penn State Worry Questionnaire, the Ruminative Response Scale, the Brief Pain Inventory, and the Pittsburgh Sleep Quality Index. Mediation analyses were performed using Hayes’ PROCESS macro (Model 4). Results: Depressive rumination was associated with pain intensity both directly (B = 0.021, 95% Confidence Intervals [CIs] 0.012, 0.030) and indirectly through sleep disturbances (B = 0.014, 95% CIs 0.010, 0.020), indicating partial mediation. In contrast, worry showed no direct effect on pain intensity (B = 0.011, 95% CIs −0.003, 0.025) but demonstrated a significant indirect effect via sleep disturbances (B = 0.018, 95% CIs 0.012, 0.025), consistent with full mediation. Conclusions: Pain intensity, sleep quality, worry, and depressive rumination are interrelated in FM. Depressive rumination plays a particularly strong role in pain perception, independent of sleep quality. Interventions that integrate cognitive–behavioral and metacognitive strategies with sleep-focused treatments may help improve both sleep and pain outcomes in individuals with FM. Full article

Background: Women living with HIV (WLHIV) in low-income urban settings face multiple intersecting nutritional risks from food insecurity, poor dietary quality, and mental health problems. We evaluated the prevalence of household food insecurity and inadequate dietary diversity, examining their associations with depressive symptoms, antiretroviral therapy (ART)-related factors, and body mass index (BMI) among WLHIV attending routine ART clinics in Kinshasa, The Democratic Republic of Congo. This study addresses critical gaps in understanding the interplay between mental health and nutrition in the context of HIV care, with significant implications for improving health outcomes among vulnerable populations. Methods: In this clinic-based cross-sectional study (February–April 2024), we enrolled 571 women on ART in Masina 2, Kinshasa. Household food insecurity was measured using the Household Food Insecurity Access Scale (HFIAS), dietary diversity was assessed using the Minimum Dietary Diversity for Women (MDD_W; inadequate ≤ 5 food groups in 24 h), and probable depression was assessed using the Hopkins Symptom Checklist-10 (HSCL-10), which is a validated screening tool. We obtained baseline BMIs from clinic records at ART induction, which we measured again upon survey completion. We used analysis of covariance (ANCOVA) to model follow-up BMI, adjusting for baseline values, age, ART duration, self-reported adherence, household food insecurity, dietary diversity, and probable depression. Sensitivity analyses included change-score and mixed-effects models. Results: The prevalence of any household food insecurity was high (75%; 95% CI:71.5–78.6), with 57.6% (95% CI:53.5–61.6) of the participants experiencing inadequate dietary diversity (MDD_W < 5). Furthermore, forty-two per cent (95% CI:38.4–46.5) experienced depressive symptoms and sixty-eight percent (95% CI: 64.4–72.0) adhered to antiretroviral therapy (ART). The mean MDD_W was 4.3, with a low consumption rate of animal-source foods. Baseline BMI was associated with follow-up values (adjusted β_{unstandardized}, 0.48 kg/m² per 1 kg/m² baseline, 95% CI 0.38–0.59; p < 0.001). Probable depression was independently associated with a lower follow-up BMI (adjusted β_{unstandardized}, −0.99 kg/m²; 95% CI −1.72 to −0.26; p = 0.008). Time since ART initiation showed a slight positive association with BMI (adjusted β_{unstandardized}, 0.10 kg/m² per year). Self-reported ART adherence, household food insecurity, and dietary diversity were not independently associated with follow-up BMI in fully adjusted models. The interaction between age and probable depression did not suggest heterogeneity between age groups (p = 0.503). Conclusions: In our cohort, food insecurity and poor dietary diversity were widespread but did not significantly correlate with BMI, while probable depression, a potentially modifiable factor, was independently associated with lower BMI after accounting for baseline nutritional status. These findings highlight the need for HIV care programs integrating mental health screening and services with nutrition-sensitive interventions to support recovery and long-term health among WLHIV. Full article

Traumatic brain injury (TBI) remains a critical public health issue worldwide, with significant morbidity, mortality, and long-term disability. Timely transfer to a specialized trauma unit is crucial to improving outcomes, yet in resource-limited settings, delays often exceed recommended time frames. This study evaluates the impact of arrival time on mortality, disability, and clinical outcomes in Ecuadorian patients with TBI. A cross-sectional and observational study was conducted, analyzing 383 adult patients diagnosed with TBI. Patients were categorized into two cohorts: those who arrived at a specialized trauma unit within five hours post-injury and those who arrived between five and 24 h. Demographic, clinical, and radiological characteristics were analyzed, including Glasgow Coma Scale (GCS), Injury Severity Score (ISS), Marshall Scale classification, and presence of subarachnoid hemorrhage (SAH). Logistic regression models were used to identify predictors of mortality and disability. Longer transfer times were associated with increased mortality (3.34 times higher for ≥5 h, p < 0.05) and disability (2.92 times higher for ≥5 h, p < 0.05). Patients with Marshall Diffuse Injury III and IV had an 8.80- and 9.05-fold increased risk of mortality, respectively. SAH was an independent predictor of mortality (4.53 times higher), and GCS between 9–13 increased the likelihood of death by 6.49 times. Delayed transfers were associated with lower GCS at admission, longer ICU stays, and increased surgical complications. Although some survivors experienced improvement over time, disability in TBI can persist for many years or even lifelong, underscoring the burden of delayed trauma care. Despite delays, overall survival remained higher than reported in high-income countries, suggesting compensatory factors in hospital-based management. Delayed hospital arrival in TBI patients significantly increases mortality and disability. Early transfer within five hours is essential to reduce secondary brain injury and improve functional outcomes. Findings suggest that in resource-limited settings, optimizing pre-hospital care and transport efficiency is crucial to minimizing long-term disability. Full article

Forests play a pivotal role in the global carbon cycle, yet accurately simulating forest soil carbon dynamics remains a significant challenge for process-based models. This review systematically compares the mechanistic foundations of traditional models (e.g., Century, CLM5) with emerging microbial-explicit models (e.g., MEND), highlighting key differences in mathematical formulation (first-order kinetics vs. Michaelis–Menten kinetics), carbon pools partitioning (measurable vs. non-measurable experimentally), and the representation of soil carbon stabilization mechanisms (inherent recalcitrance, physical protection, and chemical protection). Despite advances in process-based models in predicting forest soil organic carbon (SOC), improving prediction accuracy, and assessing SOC response to climate change, current research still faces several challenges. These include difficulties in capturing depth-dependent variations in critical microbial parameters such as microbial carbon use efficiency (CUE), limited capacity to distinguish the relative contributions of aboveground and belowground litter inputs to SOC formation, and a general lack of long-term observational data across soil profiles. To address these limitations, this study emphasizes the importance of integrating remote sensing data and refining cross-scale simulation approaches. Such improvements are essential for enhancing model predictive accuracy and establishing a more robust theoretical basis for forest carbon management and climate change mitigation. Full article

Background and Objectives: The COVID-19 pandemic has caused profound disruptions in socioeconomic, and health domains, with significant implications for mental well-being. The aim of this study was to evaluate the impact of the pandemic on stress, anxiety, and depression, alongside perceived social support, coping flexibility and related factors, and to examine how these issues vary across different age groups. Materials and Methods: A cross-sectional analytical study was conducted in Türkiye between August and December 2020. Data were collected through an online questionnaire including sociodemographic characteristics, pandemic-related concerns, and validated scales: Hospital Anxiety and Depression Scale (HADS), Perceived Stress Scale (PSS), Coping Flexibility Scale (CFS), and Multidimensional Scale of Perceived Social Support (MSPSS). Statistical analyses included descriptive and comparative tests, correlation analysis, multiple linear regression models, and correspondence analysis. Results: Among 1699 participants, 58.0% were female; 24.5% and 42.1% reported anxiety and depressive symptoms above thresholds, respectively. Younger age correlated negatively with stress, anxiety, and depression scores (p < 0.001). Feelings of loneliness, loss of control, ostracism, and sleep or concentration problems were positively associated with anxiety, depression, and stress, but negatively associated with coping flexibility and social support (p < 0.001). The 15–20 age group had the highest anxiety and depression levels and the lowest social support; the 15–30 group showed the highest stress, while the 61–75 group exhibited the lowest coping flexibility. Regression models explained 62.7% of anxiety and 56.6% of depressive symptom variances. Major predictors of anxiety included depressive symptoms, stress, and fear of dying from COVID-19, while depressive symptoms were predicted by age, stress, coping flexibility, social support, and anxiety. Conclusions: The findings highlight the considerable psychological burden and distinct vulnerabilities among age groups. Mental health interventions should be tailored according to age, emphasising the enhancement of social support and coping flexibility to strengthen resilience in future pandemics. Full article

Background: Class imbalance and limited interpretability remain major barriers to the clinical adoption of machine learning in diabetes prediction. These challenges often result in poor sensitivity to high-risk cases and reduced trust in AI-based decision support. This study addresses these limitations by integrating SMOTE-based resampling with SHAP-driven explainability, aiming to enhance both predictive performance and clinical transparency for real-world deployment. Objective: To develop and validate an interpretable machine learning framework that addresses class imbalance through advanced resampling techniques while providing clinically meaningful explanations for enhanced decision support. This study serves as a methodologically rigorous proof-of-concept, prioritizing analytical integrity over scale. While based on a computationally feasible subset of 1500 records, future work will extend to the full 100,000-patient dataset to evaluate scalability and external validity. We used the publicly available, de-identified Diabetes Prediction Dataset hosted on Kaggle, which is synthetic/derivative and not a clinically curated cohort. Accordingly, this study is framed as a methodological proof-of-concept rather than a clinically generalizable evaluation. Methods: We implemented a robust seven-stage pipeline integrating the Synthetic Minority Oversampling Technique (SMOTE) with SHapley Additive exPlanations (SHAP) to enhance model interpretability and address class imbalance. Five machine learning algorithms—Random Forest, Gradient Boosting, Support Vector Machine (SVM), Logistic Regression, and XGBoost—were comparatively evaluated on a stratified random sample of 1500 patient records drawn from the publicly available Diabetes Prediction Dataset (n = 100,000) hosted on Kaggle. To ensure methodological rigor and prevent data leakage, all preprocessing steps—including SMOTE application—were performed within the training folds of a 5-fold stratified cross-validation framework, preserving the original class distribution in each fold. Model performance was assessed using accuracy, area under the receiver operating characteristic curve (AUC), sensitivity, specificity, F1-score, and precision. Statistical significance was determined using McNemar’s test, with p-values adjusted via the Bonferroni correction to control for multiple comparisons. Results: The Random Forest-SMOTE model achieved superior performance with 96.91% accuracy (95% CI: 95.4–98.2%), AUC of 0.998, sensitivity of 99.5%, and specificity of 97.3%, significantly outperforming recent benchmarks (p < 0.001). SHAP analysis identified glucose (SHAP value: 2.34) and BMI (SHAP value: 1.87) as primary predictors, demonstrating strong clinical concordance. Feature interaction analysis revealed synergistic effects between glucose and BMI, providing actionable insights for personalized intervention strategies. Conclusions: Despite promising results, further validation of the proposed framework is required prior to any clinical deployment. At this stage, the study should be regarded as a methodological proof-of-concept rather than a clinically generalizable evaluation. Our framework successfully bridges algorithmic performance and clinical applicability. It achieved high cross-validated performance on a publicly available Kaggle dataset, with Random Forest reaching 96.9% accuracy and 0.998 AUC. These results are dataset-specific and should not be interpreted as clinical performance. External, prospective validation in real-world cohorts is required prior to any consideration of clinical deployment, particularly for personalized risk assessment in healthcare systems. Full article

Traditional engineering education struggles to bridge the theory–practice gap in the Industry 4.0 era, as static 2D schematics inadequately convey complex spatial relationships. While advanced visualization tools exist, their adoption is frequently hindered by requirements for specialized hardware and software, limiting accessibility. This study details the development and evaluation of a novel, web-based Digital Twin framework designed for accessible, intuitive e-learning that requires no client-side installation. The framework, centered on a high-fidelity 3D model of a historic radial engine, was assessed through a qualitative pilot case study with seven engineering professionals. Data was collected via a “think-aloud” protocol and a mixed-methods survey with a Likert scale and open-ended questions. Findings revealed an overwhelmingly positive reception; quantitative data showed high mean scores for usability, educational impact, and professional training potential (M > 4.2). Qualitative analysis confirmed the framework’s success in enhancing spatial understanding via features like dynamic cross-sections, improving the efficiency of accessing integrated documentation, and demonstrating high value as an onboarding tool. This work provides strong preliminary evidence that an accessible, web-based Digital Twin is a powerful and scalable solution for technical education that significantly enhances spatial comprehension and knowledge transfer. Full article

River bridge engineering alters the hydraulic characteristics of rivers, impacting fluvial morphological stability. To investigate issues concerning flood conveyance capacity within the river reach hosting a new bridge and the safe operation of existing bridges, comparative physical model tests employing both mobile-bed and fixed-bed configurations were conducted. A 1:60 scale model was used to test flood peak discharges corresponding to 30-year and 100-year return periods and investigate pier spacings of 30 m and 40 m. These tests evaluated the relative advantages and limitations of each model type in simulating flow patterns, sediment transport, and riverbed evolution. Specifically, mobile-bed models more effectively capture the interaction between water flow and sediment dynamics, while fixed-bed experiments enable more precise measurement of hydraulic parameters. Pier spacing is recognized as one of the most critical factors influencing river flow regimes. Larger pier spacing (40 m) was found to reduce upstream backwater and local scour depth compared to smaller spacing (30 m), particularly under the 30-year flood scenario. Consequently, this study investigated the effects of pier spacing on flow patterns, obtained flood conveyance characteristics under various flood frequencies, and analyzed the underlying mechanisms governing flow fields, velocity variations, and local scour around piers. The research outcomes not only elucidate multiscale coupling mechanisms between water flow and sediment but also quantify the relationship between the extent of pier-induced flow disturbance and subsequent channel morphological adjustments. This quantification provides a dynamic criterion for risk mitigation of river-crossing structures and establishes a hydrodynamic foundation for studying flood hazards in complex river reaches. Full article

Accurate bearing fault diagnosis under various operational conditions presents significant challenges, mainly due to the limited availability of labeled data and the domain mismatches across different operating environments. In this study, an adaptive meta-learning framework (AdaMETA) is proposed, which combines dynamic task-aware model-independent meta-learning (DT-MAML) with efficient multi-scale attention (EMA) modules to enhance the model’s ability to generalize and improve diagnostic performance in small-sample bearing fault diagnosis across different load scenarios. Specifically, a hierarchical encoder equipped with C-EMA is introduced to effectively capture multi-scale fault features from vibration signals, greatly improving feature extraction under constrained data conditions. Furthermore, DT-MAML dynamically adjusts the inner-loop learning rate based on task complexity, promoting efficient adaptation to diverse tasks and mitigating domain bias. Comprehensive experimental evaluations on the CWRU bearing dataset, conducted under carefully designed cross-domain scenarios, demonstrate that AdaMETA achieves superior accuracy (up to 99.26%) and robustness compared to traditional meta-learning and classical diagnostic methods. Additional ablation studies and noise interference experiments further validate the substantial contribution of the EMA module and the dynamic learning rate components. Full article

Information popularity prediction is a critical problem in social network analysis. With the increasing prevalence of social platforms, accurate prediction of the diffusion process has become increasingly important. Existing methods mainly rely on graph neural networks to model structural relationships, but they are often insufficient in capturing the complex interplay between temporal evolution and local cascade structures, especially in real-world scenarios involving sparse or rapidly changing cascades. To address this issue, we propose the Cascading Dynamic attention-calibrated Graph Convolutional Network, named CasDacGCN. It enhances prediction performance through spatiotemporal feature fusion and adaptive representation learning. The model integrates snapshot-level local encoding, global temporal modeling, cross-attention mechanisms, and a hypernetwork-based sample-wise calibration strategy, enabling flexible modeling of multi-scale diffusion patterns. Results from experiments demonstrate that the proposed model consistently surpasses existing approaches on two real-world datasets, validating its effectiveness in popularity prediction tasks. Full article

Background: NHS England has an ageing workforce. Approximately 30 percent of the NHS England workforce are aged 50 years and over, and the British Medical Association has argued that it is important that employers meet the needs of their ageing workforce and retain their skills and expertise. Objective: This sought to explore how NHS England Trusts support employees with fluctuating long-term health conditions, investigating systemic workforce challenges to providing adequate support and identifying opportunities for more inclusive and sustainable employment practices. Methods: Qualitative interviews were conducted with staff working in human resources, occupational health staff and clinical line managers involved in the support and management of staff with fluctuating long-term health conditions (n = 17). Results: The research found a misalignment between clinical managerial practices, human resource procedures, and the overarching NHS human resource policy framework, which was often seen as rigid and poorly suited to the fluctuating nature of some long-term conditions. These tensions were exacerbated by high staff turnover and limited organisational capacity. Nonetheless, instances of effective, person-centred support were also reported, typically occurring where cross-departmental collaboration and flexible, locally adapted approaches were in place. Conclusions: Findings suggest that targeted, flexible interventions for NHS employees with fluctuating long-term health conditions could enhance staff retention, reduce absenteeism, and promote more resilient workforce strategies. Identifying and scaling examples of good practice may be key to fostering a more inclusive and adaptive NHS employment model. Full article

Background: Bullying victimization is a common phenomenon that can affect middle school students’ malevolent creativity. However, the underlying mechanisms between the two remain unclear. This study integrates the social hostility model and the Conservation of Resources theory to further explore the relationship between bullying victimization and malevolent creativity, the mediating roles of trait anger and social mindfulness, and the moderating role of emotion regulation, thereby advancing the research and filling the relevant gaps. Method: Using validated Chinese versions of the Olweus Bullying Scale, Trait Anger Scale, Social Mindfulness Self-Report Scale, malevolent Creativity Behavior Scale, and Emotion Regulation Questionnaire, N = 860 students were surveyed in a cross-sectional design. Results: The results showed that bullying victimization was positively related to malevolent creativity (total effect size β = 0.44), with a direct effect of size β = 0.17 and significant indirect effects via social mindfulness (β = 0.05; 11%), trait anger (β = 0.18; 41%), and the sequential path (β= 0.04; 9%). Emotion regulation moderated the links of social mindfulness and trait anger with malevolent creativity, such that higher emotion regulation strengthened the negative association for social mindfulness and weakened the positive association for trait anger. Implications: These findings suggest that school-based programs targeting emotion regulation and social mindfulness, alongside anger management components, may help mitigate the harmful impact of bullying on malevolent creativity. Full article