Search Results (15,576)

Large flange shafts are the core load-bearing and connecting components of high-end equipment, and their welding multi-physical fields directly affect the quality and service safety of the components. Traditional experiments and finite element methods suffer from long cycles and low efficiency, which can hardly meet the demand for rapid prediction. Aiming at the fast and accurate prediction of welding temperature, deformation and residual stress, this study combines thermal–mechanical coupled finite element simulation with machine learning to construct and compare a variety of prediction models. A dataset is built based on simulation data from 100 groups of process parameters. Overfitting is reduced through strategies including early stopping and dropout, and models such as MLP, RF, RBF-SVR, TabNet, XGBoost, and FT-Transformer are established and verified through 10-fold cross-validation. The results show that the MLP model performs best in the prediction of temperature, deformation and residual stress, and is in good agreement with the simulation values. The prediction errors of the peak temperature of the weld and base metal are below 5%, and the errors of deformation and residual stress are controlled within 10%. The average error of peak residual stress is about 6 MPa, and the deviation of most samples is less than 5 MPa. The RF model ranks second in accuracy, with an average error of about 6.5 MPa for peak residual stress, showing a satisfactory interpretability and engineering applicability. RBF-SVR and TabNet can meet basic prediction requirements. Under the small-sample condition in this work, XGBoost and FT-Transformer present relatively large errors and a weak generalization ability, making it difficult to achieve high-precision prediction. The MLP model established in this paper can effectively reproduce the evolution of welding multi-physical fields and supports the rapid prediction and process optimization of large flange shaft welding. The generalization ability and practical performance of the model can be further improved by expanding the dataset and experimental verification in the future. Full article

Flood change detection from remote sensing data can be used to identify post-disaster flooded areas, providing decision support for emergency rescue and post-disaster reconstruction. Although the combination of SAR and optical images effectively addresses obscuration by clouds and rain, the inherent difference in their imaging mechanisms poses a challenge to improving the accuracy of flood area change detection. Furthermore, existing flood inundation change detection methods based on heterogeneous remote sensing imagery struggle to distinguish small ground objects within the background from the actual inundated regions. Therefore, a pseudo-Siamese hybrid dilation spiral feature network (PSiam-HDSFNet) is proposed in this paper. Firstly, the feature extraction pipeline progressively processes optical and SAR images through five-layer Enhanced Deep Residual Blocks and five-layer Residual Dense Blocks, respectively. A Hybrid Dilated Pyramid (HDP) module based on a sawtooth wave-like dilated coefficient is designed to enhance multi-scale semantics of deep features in order to selectively reinforce semantic features in flood areas and weaken the noise semantics from small ground objects. Then, a Spiral Feature Pyramid (SFP) module is designed to make the deep features of SAR and optical images more consistent in spatial structure and numerical distribution patterns, so that the features of flood areas become more prominent while the noise semantics from small ground objects are further suppressed. After that, the Galerkin-type attention with linear complexity is introduced to the decoder, rapidly reconstructing the abstract semantic information of floods into interpretable flood features. Finally, the Align OPT-SAR (AlignOS) method is designed to align SAR and optical image features, enabling subsequent flood area detection. Seven metrics are adopted in the comparison between PSiam-HDSFNet and the other 14 methods. The results indicate that PSiam-HDSFNet improves change detection accuracy by extracting and processing depth features of these two images without image domain translation, and its F1 scores are improved by 7.704%, 7.664%, 4.353%, and 1.111% in the four flood coverage categories detection tasks compared to the suboptimum. Full article

This study examined the role of e-modules and learning applications in enhancing scientific literacy among Grade 11 Life Sciences learners in a South African secondary school. Grounded in constructivist and connectivist learning theories, the research responded to persistent challenges in learners’ conceptual understanding, scientific reasoning, and application of scientific knowledge. A mixed-methods case study design was employed, combining quantitative pre- and post-test data with qualitative classroom observations and semi-structured learner interviews. Thirty learners participated in a technology-enhanced instructional intervention using curriculum-aligned e-modules delivered through Binogi and Google Classroom. Quantitative findings revealed a statistically significant improvement in scientific literacy following the intervention. Learners’ mean scores increased from 39.20% (pre-test) to 63.07% (post-test), representing a gain of 23.87 percentage points. A paired-samples t-test confirmed that this improvement was extremely significant (t (29) = 11.58, p < 0.0001), with a very large effect size (Cohen’s d = 2.11). Qualitative findings indicated that learners experienced enhanced engagement, improved conceptual clarity, and greater motivation when using digital learning tools, particularly through visualisations, animations, and self-paced learning. However, persistent difficulties with graph interpretation were also identified. The study concludes that the intentional integration of e-modules and learning applications can substantially enhance scientific literacy in Life Sciences by supporting conceptual understanding, reasoning, and learner engagement. These findings highlight the importance of pedagogically guided digital integration and teacher professional development to strengthen science learning outcomes. Full article

Background/Objectives: There is little information on the definition and management of dyslipidemia in patients and survivors of childhood, adolescent and young adult (CAYA) cancer. However, it is known that this population is at higher risk of developing cardiovascular disease (CVD). Dyslipidemia, hypertension, and metabolic syndrome are common among CAYA cancer survivors due to the cancer itself or the treatment received. Therefore, managing dyslipidemia in this population is crucial to mitigate the risk of long-term CVD. The aim of this systematic review was to summarize currently used definitions and cutoffs for dyslipidemia, its prevalence, and management strategies in CAYA cancer survivors. This review further describes reported pharmacological and lifestyle interventions and their impact on lipid levels. Methods: We conducted a systematic literature search in PubMed, including studies published in English, German or French between January 2015 and February 2025. The population included individuals diagnosed with any type of CAYA cancer prior to 25 years of age and either receiving cancer treatment or in follow-up care. We considered all types of publications except for Phase I and II studies. We followed PRISMA guidelines, assessed the quality of the eligible studies according to the respective Joanna Briggs Institute’s Critical Appraisal Tools, and reported the results descriptively. Results: Of 575 identified publications, 53 fulfilled the inclusion criteria. Forty-three studies reported on the definitions of abnormal lipid values, 40 stated the prevalence of abnormal lipid values, and 17 studies described management approaches, of which 12 were case reports and small case series. For all three outcomes, the results were very heterogeneous. Using the example of triglycerides (TGs), the cutoff values for high TGs ranged from 5.17 mmol/L to 6.2 mmol/L and the reported prevalence of high TGs ranged from 0% to 75%, with an average of 31%. The only reported intervention to prevent dyslipidemia in CAYA cancer survivors was lifestyle modification. Preventive strategies that started during treatment were lifestyle modifications and fish oil supplements. Conclusions: Our systematic review provides a comprehensive overview of existing definitions, prevalences, and management of abnormal lipid values in CAYA cancer patients and survivors. However, the identified heterogeneities indicate that reported prevalences and results of interventions must be interpreted cautiously. An internationally harmonized approach to defining and reporting lipid values in CAYA cancer survivors is urgently needed to enable tailored screening and treatment strategies. Full article

Background: Adequate iodine intake during pregnancy is essential for optimal maternal thyroid function and fetal neurodevelopment. Although universal salt iodization has been implemented in Turkey, pregnant women may remain vulnerable to iodine insufficiency. This study aimed to evaluate maternal thyroid function in relation to iodine status, and to comprehensively compare the diagnostic performance of spot urinary indices and creatinine-adjusted measures against measured 24 h urinary iodine excretion (24h-UIE) in pregnant women. Methods: A total of 227 pregnant women attending family health centers in Afyonkarahisar, Turkey, provided both spot urine samples and complete 24 h urine collections. Urinary iodine concentration (UIC), creatinine-corrected UIC (UIC/UCr), and 24h-UIE were measured. Thyroid function tests were interpreted using trimester-specific reference ranges. Correlations between urinary indices were assessed, and ROC analyses were performed using 24h-UIE as the operational reference. A structured questionnaire evaluated iodine-related dietary knowledge and salt-use practices. Results: The median spot UIC was 59.0 µg/L, indicating insufficient recent iodine intake at the population level. Based on 24h-UIE, 70% of participants had excretion levels below the Estimated Average Requirement (EAR) threshold (<144 µg/day). Spot UIC showed a weak correlation with 24h-UIE (rho = 0.270, p < 0.001), whereas UIC/UCr demonstrated a stronger correlation (rho = 0.491, p < 0.001). In ROC analyses, UIC/UCr yielded a significantly higher AUC than spot UIC (0.774 [95% CI: 0.707–0.841] vs. 0.670 [95% CI: 0.593–0.748]; DeLong p = 0.016). Overt hypothyroidism was not observed; subclinical hypothyroidism was present in 16.3% of participants. While no overall association was found between iodine indices and thyroid status, in the first trimester, those with subclinical hypothyroidism had higher 24h-UIE medians than euthyroid peers (134.2 vs. 100.3 µg/day, p = 0.037), although both groups remained below the EAR threshold. Knowledge regarding iodine-rich foods and iodized salt use was limited among the study population. Conclusions: Iodine insufficiency remains highly prevalent among pregnant women in this region despite salt iodization. While spot UIC alone showed limited agreement with 24h-UIE, creatinine-adjusted UIC may offer improved interpretability under conditions of variable urine dilution. Preserved thyroid function in the presence of iodine insufficiency highlights the silent nature of this condition during pregnancy. Strengthened pregnancy-specific iodine surveillance and targeted antenatal education are warranted. Full article

Stress can cause many disorders, including mental and physical ones, if it persists. To take timely and effective early intervention measures, mental stress levels must be carefully monitored. This study investigates the rapidly growing topic of mental stress detection, focusing on the primary goals and mechanisms of existing detection frameworks. The main objectives and mechanisms will be highlighted. This study examines physiological sensors, stressors, algorithms, monitoring methods, and validation tools used to assess and classify mental stress. The study targets physiological sensors. Wearable sensors are becoming more popular because they can continuously monitor physiological responses in human-like environments. This allows them to reveal relevant stress patterns across various work environments. Numerous physiological sensors are used regularly. Galvanic skin response (GSR), electrocardiogram (ECG), photoplethysmography (PPG), electroencephalography (EEG), and pupil diameter camera systems are examples of these sensors. The combination of these sensors provides a wealth of cognitive and autonomic response data for stress detection. This review examines AI-based methods for interpreting complex physiological data. Machine learning and ensemble models are emphasized for improving stress classification accuracy and reducing incorrect classifications. In addition, this article discusses stressors used to induce reliable physiological responses. Validated self-report instruments are being reviewed as benchmarking tools for objective sensor-based measurements. STAI and PSS-10 are examples. These instruments demonstrate a strong correlation between stress and anxiety and physiological health outcomes. In conclusion, this review discusses future research avenues, focusing on advanced artificial intelligence-driven approaches and sophisticated sensors. These developments aim to better define stress levels and physiological factors that have not been thoroughly studied. Full article

Background/Objectives: Co-testing may improve cervical cancer prevention by stratifying women into groups with different absolute risks of CIN2+, CIN3+, and cervical cancer. We evaluated real-world co-testing with cervical cytology and a genotype-specific HPV E6/E7 mRNA assay targeting HPV16, HPV18, and HPV45 (PreTect SEE) in routine clinical practice. Methods: We conducted a retrospective, registry-based cohort study at the Department of Clinical Pathology, University Hospital of North Norway. Eligible co-test samples (liquid-based cytology with concurrent HPV mRNA testing, both with valid results) from routine screening, follow-up, and clinically indicated testing were identified from the laboratory information system and passively followed for worst histological outcome through December 2025. Outcomes were no biopsy/<CIN2, CIN2+, CIN3+, and cervical cancer. Results: Among 116,217 eligible co-test samples (mean age 43.9 years), cumulative risks were 4.4% for CIN2+, 1.5% for CIN3+, and 0.1% for cervical cancer. Baseline HPV mRNA positivity was 3.9%, and cytology was ASC-US+ in 12.2% of samples. Co-testing produced a clear stepwise risk gradient. Double-negative results (NILM/mRNA−; 86.7%) had very low risks (CIN3+ 0.2%; cervical cancer 0.02%). ASC-US+/mRNA− results (9.3%) showed intermediate risks (CIN3+ 4.1%; cervical cancer 0.2%). NILM/mRNA+ results (1.1%) showed substantially higher risks despite normal cytology (CIN3+ 13.0%; cervical cancer 0.5%). Double-positive results (ASC-US+/mRNA+; 2.8%) had the highest risks (CIN3+ 28.5%; cervical cancer 2.3%). Within NILM, mRNA positivity captured 42.7% of CIN3+ cases and 25.0% of cancers. Genotype-specific analyses showed highest risks for HPV16, followed by HPV18 and HPV45. Conclusions: Co-testing with cervical cytology and a 3-type HPV mRNA assay provided strong, clinically interpretable risk stratification and identified a small but high-risk subgroup among women with normal cytology. These findings support genotype-specific HPV mRNA testing as an adjunct to cytology in routine care. Full article

This study presents an explainable knowledge graph (KG) framework that transforms continuous spindle monitoring time-series data into transparent, reasoning-ready diagnostic structures. Existing data-driven approaches, while accurate, often lack the interpretability required for high-stakes industrial decision-making and are sensitive to operating condition drifts. To address these limitations, we propose a two-level temporal segmentation method combining label transition detection and statistical drift analysis to identify meaningful state boundaries. Furthermore, a percentile-based discretization mechanism converts statistical features into interpretable semantic tags. A Neo4j-based state–event–feature schema captures lifecycle evolution and evidence relations, enabling attribution path reasoning that links failure events to salient precursor features. Experiments on real industrial spindle data demonstrate a fault detection accuracy of 84.97% and a false alarm rate of 3.43%, effectively capturing stable baselines and intermittent abnormal bursts. The proposed framework provides a distinct novelty in bridging the gap between numerical time-series and symbolic reasoning, offering a practical pathway for deploying explainable and maintainable spindle health analytics. Full article

Background/Objectives: Fibromyalgia is a chronic pain syndrome influenced by both physical and psychological factors, but their interaction remains unclear. We evaluated tools combining physical and emotional dimensions to characterise fibromyalgia and assess associations with persistent physical symptoms (PPS) and the emotional distress in its clinical interpretation. Methods: This cross-sectional study included 1588 patients referred to the Pain Management Unit, Complejo Asistencial Universitario de León. Fibromyalgia cases had a prior diagnosis to referral using the 2019 ACTTION–American Pain Society Pain Taxonomy diagnostic criteria. At the first consultation, participants completed a standardized protocol including sociodemographic variables, a Central Sensitization Inventory Part B checklist of previously physician-diagnosed physical and psychological conditions, and the visual analogue scale for pain, the Modified Somatic Perception Questionnaire (MSPQ), and the Zung Self-Rating Depression Scale. The Distress and Risk Assessment Method (DRAM) was used to integrate MSPQ and Zung data. Results: Women had higher odds of fibromyalgia (p = 0.003). Fibromyalgia was associated with PPS (p < 0.001), with chronic fatigue predominating in women (p < 0.001) and neck injury/whiplash in men (p = 0.005). The MSPQ had the highest OR among the instruments evaluated (overall: p < 0.001; women: p < 0.001; men: p = 0.005). Fibromyalgia status differed by DRAM category (nominal model, p < 0.001), suggesting higher odds in the Depressive and Somatic categories compared with Normal. Conclusions: In our sample, sex was associated with fibromyalgia and PPS profiles; PPS profiles were also associated with fibromyalgia. The MSPQ appeared to be among the most informative instruments, and its integration with Zung through the DRAM may have potential utility for psychosomatic risk profiling in fibromyalgia, warranting further study. Full article

Source camera identification on smartphones constitutes a fundamental task in multimedia forensics, providing essential support for applications such as image copyright protection, illegal content tracking, and digital evidence verification. Numerous techniques have been developed for this task over the past decades. Among existing approaches, Photo-Response Non-Uniformity (PRNU) has been widely recognized as a reliable device-specific fingerprint and has demonstrated remarkable performance in real-world applications. Nevertheless, the rapid advancement of computational photography technologies has introduced significant challenges: modern devices often exhibit anomalous behaviors under PRNU-based analysis. For instance, images captured by different devices may exhibit unexpected correlations, while images captured by the same device can vary substantially in their PRNU patterns. Current approaches are incapable of automatically exploring the underlying causes of these anomalous behaviors. To address this limitation, we propose a simple yet effective forensic analysis framework leveraging Exchangeable Image File Format (EXIF) metadata. Specifically, we represent EXIF metadata as type-aware word embeddings to preserve contextual information across tags. This design enables visual interpretation of the model’s decision-making process and provides complementary insights for identifying the anomalous behaviors observed in modern devices. Extensive experiments conducted on three public benchmark datasets demonstrate that the proposed method not only achieves state-of-the-art performance for source camera identification but also provides valuable insights into anomalous device behaviors. Full article

Background/Objectives: Fundus imaging provides a detailed view of the interior surface of the eye and plays a crucial role in the early diagnosis of retinal diseases. However, automated interpretation of fundus images remains challenging due to variations in illumination, noise, and structural complexity. Methods: A novel hybrid model that integrates the Intrinsic Mode Function (IMF) filter, derived from Empirical Mode Decomposition (EMD), with a Light Convolutional Neural Network (LightCNN) for enhanced fundus image classification was proposed. The IMF filter effectively decomposes the input signal into intrinsic components, isolating high-frequency noise and preserving critical retinal patterns. These refined components are subsequently processed by the LightCNN architecture, which offers lightweight yet highly discriminative feature extraction and classification capabilities. Results: Experimental results on DIARETDB fundus datasets demonstrate that the proposed IMF + LightCNN model achieves 99.4% accuracy, 99.1% precision, 98.87% recall, and a 98.31 F1-score, significantly outperforming conventional CNN and ResNet-based models. Conclusions: Integrating advanced signal processing with lightweight deep learning improves both diagnostic accuracy and computational efficiency. This hybrid framework establishes a promising pathway for reliable and real-time clinical screening of retinal diseases. Full article

Background: Dental radiographs are essential for diagnosis and treatment planning in modern dentistry. However, their manual interpretation is time-consuming and subject to variability, highlighting the need for automated tools to improve efficiency and consistency. This study aims to validate ORTHOSEG, a deep learning-based system designed to automate the segmentation of anatomical, pathological, and non-pathological elements in radiographs, including orthopantomograms, bitewings, and periapical images. Methods: ORTHOSEG’s performance was evaluated using a rigorously curated dataset of 150 dental radiographs, including 50 orthopantomograms, 50 bitewings, and 50 periapical images, with manual annotations by expert clinicians serving as the ground truth. The system’s segmentation performance was assessed using standard evaluation metrics, including mean Dice Similarity Coefficient (mDSC) and mean Intersection over Union (mIoU), and inference time was also recorded. Results: The system achieved high accuracy, with mDSC and mIoU values of 0.635 ± 0.233 and 0.576 ± 0.214, respectively. In particular for orthopantomograms, it achieved an mDSC of 0.756 ± 0.174 and an mIoU of 0.684 ± 0.172, surpassing existing benchmarks. Its segmentation capabilities extend to approximately 70 distinct elements, underscoring its comprehensive utility. The system demonstrated efficient computational performance, with processing times of 19.745 ± 3.625 s for orthopantomograms, 8.467 ± 0.903 s for bitewings, and 5.653 ± 0.897 s for periapical radiographs on standard clinical hardware. Conclusions: ORTHOSEG demonstrates efficiency suitable for integration into routine workflows. This study confirms ORTHOSEG’s reliability and potential to improve diagnostic workflows, offering clinicians a valuable tool for faster and more detailed radiograph analysis. Future research will focus on extending validation across diverse clinical scenarios to ensure broader applicability. However, this study has limitations, including the use of a dataset derived from a European population and the absence of usability and clinical workflow evaluation, which should be addressed in future studies. Full article

The geothermal system in the Jiaodong Peninsula is situated within a continent–ocean transition zone, where complex interactions among meteoric water, geothermal fluids, and seawater produce diverse hydrogeochemical and isotopic signatures, complicating geothermal resource assessment and sustainable development. To constrain recharge sources and seawater mixing mechanisms, geothermal water samples were systematically collected from 15 geothermal fields and analyzed using integrated hydrogeochemical methods and multi-isotope tracers (δD–δ¹⁸O, δ³⁴S-SO₄²⁻, ⁸⁷Sr/⁸⁶Sr, and ³H). The results show that geothermal waters are predominantly recharged by meteoric precipitation, with δD–δ¹⁸O values distributed along the meteoric water line, while low d-excess values indicate prolonged circulation and significant water–rock interaction. Seawater mixing exhibits marked spatial heterogeneity: only 5 of the 15 fields show detectable marine influence. Chloride-based calculations suggest apparent seawater fractions of up to ~34% in BQ and <4% in DY, whereas the remaining fields show negligible mixing. Sulfur and strontium isotopes indicate contributions from external sulfate sources and continued water–rock interaction rather than simple mixing with modern seawater. Low tritium contents further imply involvement of deeply circulated paleo-seawater. The system is therefore interpreted as a fault-controlled seawater-mixing geothermal system, providing insights into coastal geothermal evolution and resource evaluation. Full article

This paper investigates a fractional-order mathematical model for the co-infection dynamics of pneumonia and typhoid fever using the Liouville–Caputo derivative. We establish the existence, uniqueness, non-negativity, and boundedness of solutions using Banach’s fixed point theorem and fractional comparison principles. The Hyers–Ulam and generalized Ulam–Hyers–Rassias stability of the system are rigorously proved; this stability analysis is epidemiologically significant because it guarantees that small perturbations in initial conditions or model parameters—inevitable in real-world data collection—do not lead to unbounded deviations in disease trajectory predictions. To approximate solutions numerically, we develop a Laplace-Based Optimized Decomposition Method (LODM) and validate its convergence against a modified predictor–corrector scheme. The LODM provides a semi-analytical series solution, while the predictor–corrector method serves as a numerical benchmark; this dual approach ensures reliability of simulations. Numerical simulations illustrate the influence of the fractional order $\xi$ on system dynamics. Quantitative comparison between $\xi =1$ (integer order) and $\xi <1$ (fractional order) demonstrates that fractional modeling reduces peak infection by 12–18% and delays epidemic peaks by 15–30 days, confirming that memory effects capture long-term epidemiological dependencies that integer-order models fail to reproduce. A biological interpretation links the fractional order to immune memory, pathogen persistence, and intervention latency. This study provides both theoretical and numerical evidence supporting the use of fractional calculus in epidemiological modeling. Full article

Background/Objectives: Head and neck adenoid cystic carcinoma (HNAdCC) is characterized by indolent growth but sustained long-term risk of late recurrence and disease-related mortality. Data describing very long-term outcomes using analytic approaches that explicitly account for competing mortality remain limited. We aimed to characterize late failures, competing causes of death, and clinically interpretable long-horizon risk estimates after primary surgery for HNAdCC. Methods: We performed a retrospective single-center cohort study of patients with HNAdCC treated with curative-intent surgery between 1984 and 2020. Overall survival (OS) and cancer-specific survival (CSS) were estimated using Kaplan–Meier method. Competing risks of disease-related and other-cause death, as well as first-failure patterns, were analyzed using cumulative incidence functions, including a 5-year landmark analysis. Conditional mortality and restricted mean survival time (RMST; τ = 25 years) were additionally assessed. Results: Fifty-seven patients were included (median age 54 years). Median follow-up was 133 months overall and 212 months among survivors. A first failure occurred in 19/57 (33.3%) of patients, with distant metastasis as the most common pattern; 7/19 (36.8%) of failures occurred beyond 5 years. OS at 5, 10, and 25 years was 68.4%, 64.9%, and 37.5%, respectively; corresponding CSS was 78.9%, 74.8%, and 51.7%. At 25 years, cumulative incidence of disease-related death was 41.7%, compared with 20.9% for other-cause death. Older age and advanced T category were independently associated with worse OS, while older age and perineural invasion predicted worse CSS. Among 5-year survivors, conditional risk of disease-related death by 25 years remained 32.7%. RMST analyses demonstrated substantial long-term life-years lost associated with perineural invasion and T3–4 disease. Conclusions: HNAdCC exhibits persistent long-term risk with clinically meaningful late failures and substantial competing mortality over decades. Conditional and RMST-based estimates provide patient-centered measures that support lifelong, risk-adapted surveillance, particularly focused on detection of distant metastases. Full article