Search Results (1,745)

Background: Chest X-ray (CXR) is widely used for the assessment of thoracic diseases, yet automated multi-label interpretation remains challenging due to subtle visual patterns, overlapping anatomical structures, and frequent co-occurrence of abnormalities. While recent deep learning models have shown strong performance, limitations in interpretability, anatomical awareness, and robustness continue to hinder their clinical adoption. Methods: The proposed framework employs a hybrid ConvNeXtV2–Vision Transformer (ViT) architecture that combines convolutional feature extraction for capturing fine-grained local patterns with transformer-based global reasoning to model long-range contextual dependencies. The model is trained exclusively using image-level annotations. In addition to classification, three complementary post hoc components are integrated to enhance model trust and interpretability. A segmentation-aware Gradient-weighted class activation mapping (Grad-CAM) module leverages CheXmask lung and heart segmentations to highlight anatomically relevant regions and quantify predictive evidence inside and outside the lungs. An ontology-driven neuro-symbolic reasoning layer translates Grad-CAM activations into structured, rule-based explanations aligned with clinical concepts such as “basal effusion” and “enlarged cardiac silhouette”. Furthermore, a lightweight out-of-distribution (OOD) detection module based on confidence scores, energy scores, and Mahalanobis distance scores is employed to identify inputs that deviate from the training distribution. Results: On the VinBigData test set, the model achieved a macro-AUROC of 0.9525 and a Micro AUROC of 0.9777 when trained solely with image-level annotations. External evaluation further demonstrated strong generalisation, yielding macro-AUROC scores of 0.9106 on NIH ChestXray14 and 0.8487 on CheXpert (frontal views). Both Grad-CAM visualisations and ontology-based reasoning remained coherent on unseen data, while the OOD module successfully flagged non-thoracic images. Conclusions: Overall, the proposed approach demonstrates that hybrid convolutional neural network (CNN)–vision transformer (ViT) architectures, combined with anatomy-aware explainability and symbolic reasoning, can support automated chest X-ray diagnosis in a manner that is accurate, transparent, and safety-aware. Full article

This paper presents the simultaneous maximization of speed and sensitivity in the Optimal Coordination of Directional Over-Current Protections (OC-DOCP), considering that maximum selectivity is maintained in all solutions. Only these three desirable features of the protection system were considered in the multi-objective approach; thus, the problem can be simply formulated as the weighted sum of speed and sensitivity as goals to be maximized, and the Pareto frontiers correlating speed and sensitivity are easily found in this way. These Pareto frontiers had not been shown in the literature about this topic, and they properly show the compromise solutions for the optimal solutions (i.e., speed improvements imply sensitivity deterioration while sensitivity improvements imply speed degradation). The simplest OC-DOCP formulation, applied to a well-known sample system, is taken as an example to show the Pareto frontiers for different time–current curve types. Another OC-DOCP formulation, which considers different topologies and their probability of occurrence, is also solved and the corresponding Pareto frontiers are also shown. The main findings of this work are the following: (a) in general, the results show that the variation in the speed in the Pareto frontier is more notorious for the less inverse curve types, whose optimal solutions are slower; (b) in the case of extremely inverse curves, the optimal solutions are faster and the effect of changes in sensitivity on the protection speed is very low in the Pareto frontiers; (c) it is also herein shown that the knowledge of this topic is also useful to solve some possible cases of unfeasibility related to the upper bound of time dial settings. Full article

The causes of agricultural drought are complex, and its actual occurrence process is often characterized by rapid onset in terms of time and small scale in terms of space. Monitoring agricultural drought using satellite remote sensing with low spatial resolution makes it difficult to accurately capture the details of small-scale drought events. High-resolution satellite remote sensing has relatively long revisit cycles, making it difficult to capture the rapid evolution of drought conditions. Furthermore, the occurrence of agricultural drought is linked to multiple factors including precipitation, evapotranspiration, soil properties, and crop physiological characteristics. Consequently, relying on a single variable or indicator is insufficient for multidimensional monitoring of agricultural drought. This study takes Hebi City, Henan Province as the research area. It uses Sentinel-1 satellite data (HV, VV), Sentinel-2 data (NDVI, B2, B11), elevation, slope, aspect, and GPM precipitation data from 2019 to 2024 as independent variables. Three machine learning algorithms—Random Forest (RF), Random Forest-Recursive Feature Elimination (RF-RFE), and eXtreme Gradient Boosting (XGBoost)—were employed to construct a multi-dimensional agricultural drought monitoring model at the field scale. Additionally, the study verified the sensitivity of different environmental variables to agricultural drought monitoring and analyzed the accuracy performance of different machine learning algorithms in agricultural drought monitoring. The research results indicate that under the condition of full-factor input, all three models exhibit the optimal predictive performance. Among them, the XGBoost model performs the best, with the smallest Relative Root Mean Square Error (RRMSE) of 0.45 and the highest Correlation Coefficient (R) of 0.79. The absence of Digital Elevation Model (DEM) data impairs the models’ ability to capture the patterns of key features, which in turn leads to a reduction in predictive accuracy. Meanwhile, there is a significant correlation between model performance and sample size. Ultimately, the constructed XGBoost model takes the lead with an accuracy of 89%, while the accuracies of Random Forest (RF) and Random Forest-Recursive Feature Elimination (RF-RFE) are 88% and 86%, respectively. Based on these three drought monitoring models, this study further monitored a drought event that occurred in Hebi City in 2023, presented the spatiotemporal distribution of agricultural drought in Hebi City, and applied the Mann–Kendall test for time series analysis, aiming to identify the abrupt change process of agricultural drought. Meanwhile, on the basis of the research results, the feasibility of verifying drought occurrence using irrigation signals was discussed, and the potential reasons for the significantly lower drought occurrence probability in the western mountainous areas of the study region were analyzed. Full article

The family Capitellidae performs critical roles in bioturbation and sediment remediation within global marine benthic ecosystems. However, they are a taxonomically challenging group due to their simple morphology and a ‘morphological mosaic’, where traditional classificatory traits, such as thoracic chaetiger counts, appear convergently across genera. Previous multi-locus studies (using 18S, 28S, H3, and COI) first highlighted this conflict, revealing the polyphyly of major genera like Notomastus and even Leiochrides itself (based on unidentified specimens). More recently, mitogenomic studies uncovered massive gene order rearrangements and a conflicting topology but did not include Leiochrides. Critically, with no complete mitogenome reported for a formally identified Leiochrides species, its true phylogenetic position and the validity of its polyphyly remain unresolved. To address this critical gap, we sequenced and characterized the first complete mitochondrial genome from a formally identified species, Leiochrides yokjidoensis, recently described from Korean waters. The complete mitogenome was 17,933 bp in length and included the typical 13 protein-coding genes (PCGs), 2 ribosomal RNAs (rRNAs), and 22 transfer RNAs (tRNAs). Gene order (GO) analysis revealed the occurrence of gene rearrangements in Capitellidae and in its sister clade, Opheliidae. A phylogenomic analysis using the amino acid sequences of 13 PCGs from 30 species established the first robust systematic position for the genus Leiochrides (based on this formally identified species). Phylogenetic results recovered Leiochrides as a sister group to the clade comprising Mediomastus, Barantolla, Heteromastus, and Notomastus hemipodus (BS 99%). This distinct placement confirms that Leiochrides represents an independent evolutionary lineage, phylogenetically separate from the polyphyletic Notomastus complex, despite their morphological similarities. Furthermore, our analysis confirmed the polyphyly of Notomastus, with N. hemipodus clustering distinctly from other Notomastus species. Additionally, signatures of positive selection were detected in ND4, and ND5 genes, suggesting potential adaptive evolution to the subtidal environment. This placement provides a critical, high-confidence anchor point for the genus Leiochrides. It provides a reliable reference to investigate the unresolved polyphyly suggested by previous multi-locus studies and provides compelling evidence for the hypothesis that thoracic chaetiger counts are of limited value for inferring phylogenetic relationships. This study provides the foundational genomic cornerstone for Leiochrides, representing an essential first step toward resolving the systematics of this taxonomically challenging family. Full article

Redox (reduction–oxidation) processes underlie all forms of life and are a universal regulatory mechanism that maintains homeostasis and adapts the organism to changes in the internal and external environments. From capturing solar energy in photosynthesis and oxygen generation to fine-tuning cellular metabolism, redox reactions are key determinants of life activity. Proteins containing sulfur- and selenium-containing amino acid residues play a crucial role in redox regulation. Their reversible oxidation by physiological oxidants, such as hydrogen peroxide (H₂O₂), plays the role of molecular switches that control enzymatic activity, protein structure, and signaling cascades. This enables rapid and flexible cellular responses to a wide range of stimuli—from growth factors and nutrient signals to toxins and stressors. Mitochondria, the main energy organelles and also the major sources of reactive oxygen species (ROS), play a special role in redox balance. On the one hand, mitochondrial ROS function as signaling molecules, regulating cellular processes, including proliferation, apoptosis, and immune response, while, on the other hand, their excessive accumulation leads to oxidative stress, damage to biomolecules, and the development of pathological processes. So, mitochondria act not only as a “generator” of redox signals but also as a central link in maintaining cellular and systemic redox homeostasis. Redox signaling forms a multi-layered cybernetic system, which includes signal perception, activation of signaling pathways, the initiation of physiological responses, and feedback regulatory mechanisms. At the molecular level, this is manifested by changes in the activity of redox-regulated proteins of which the redox proteome consists, thereby affecting the epigenetic landscape and gene expression. Physiological processes at all levels of biological organization—from subcellular to systemic—are controlled by redox mechanisms. Studying these processes opens a way to understanding the universal principles of life activity and identifying the biochemical mechanisms whose disruption causes the occurrence and development of pathological reactions. It is important to emphasize that new approaches to redox balance modulation are now actively developed, ranging from antioxidant therapy and targeted intervention on mitochondria to pharmacological and nutraceutical regulation of signaling pathways. This article analyzes the pivotal role of redox balance and its regulation at various levels of living organisms—from molecular and cellular to tissue, organ, and organismal levels—with a special emphasis on the role of mitochondria and modern strategies for influencing redox homeostasis. Full article

The rapid expansion of wind power as a key component of global renewable energy systems has led to the widespread deployment of wind turbines in environments exposed to diverse natural hazards. While hazard effects are often investigated individually, real wind turbine systems frequently experience concurrent or sequential hazards over their operational lifetime, giving rise to interaction effects that are not adequately captured by conventional design approaches. This paper presents Part 2 of a comprehensive review on natural hazards affecting wind turbine performance, combining bibliometric keyword co-occurrence analysis with a critical synthesis of recent technical studies. The review focuses on earthquakes, sea waves, and extreme wind events, while also highlighting other hazard types that have received comparatively limited attention in the literature, examining their effects on wind turbine systems and the mitigation strategies reported to address associated risks. Rather than treating hazards in isolation, their impacts are synthesised through cross-hazard interaction pathways and component-level failure modes. The findings indicate that wind turbine vulnerability under multi-hazard conditions is governed not only by load magnitude but also by hazard-induced changes in system properties and operational state. Key research gaps are identified, emphasising the need for state-aware, mechanism-consistent multi-hazard assessment frameworks to support the resilient design and operation of future wind energy systems. Full article

Many nations have been investing in hydrogen energy in the most recent wave of development and numerous projects have been proposed, yet a substantial share of these projects remain at the conceptual or feasibility stage and have not progressed to final investment decision or operation. There is a need to identify initial potential sites for green hydrogen production from renewable energy on an objective basis with minimal upfront cost to the investor. This study develops a decision support system (DSS) for identifying optimal locations for green hydrogen production using solar and wind resources that integrate economic, environmental, technical, social, and risk and safety factors through advanced Multi-Criteria Decision Making (MCDM) techniques. The study evaluates alternative weighting scenarios using (a) occurrence-based, (b) PageRank-based, and (c) equal weighting approaches to minimize human bias and enhance decision transparency. In the occurrence-based approach (a), renewable resource potential receives the highest weighting (≈34% total weighting). By comparison, approach (b) redistributes importance toward infrastructure and social indicators, yielding a more balanced representation of technical and economic priorities and highlighting the practical value of capturing interdependencies among indicators for resource-efficient site selection. The research also contrasts the empirical and operational efficiencies of various weighting methods and processing stages, highlighting strengths and weaknesses in supporting sustainable and economically viable site selection. Ultimately, this research contributes significantly to both academic and practical implementations in the green hydrogen sector, providing a strategic, data-driven approach to support sustainable energy transitions. Full article

The MALDI-TOF MS Bruker Biotyper MBT subtyping IVD module enables the early detection of cfiA-positive Bacteroides fragilis (cfiA+ BF) during bacterial identification. However, the relationship between genetic positivity, phenotypic resistance, and clinical outcomes has not been fully elucidated. This retrospective study analyzed B. fragilis isolates from three Hong Kong hospitals between 2021 and 2025 to examine their prevalence and the clinical utility of MALDI-TOF MS in rapid cfiA detection. Antibiotic susceptibility testing, cfiA gene detection using MALDI-TOF MS, and Oxford Nanopore sequencing were performed. Medical records were reviewed, and univariate analyses and multivariate logistic regression were used to identify factors associated with cfiA positivity and 30-day all-cause mortality. Overall, B. fragilis exhibited a high rate of antibiotic resistance. Concomitant resistance to carbapenems and metronidazole was identified in three isolates. Among the 166 isolates, 40 (24.1%) were cfiA-positive. cfiA detection by MALDI-TOF MS showed 100% concordance with the gene sequencing results and correlated strongly with phenotypic carbapenem resistance (Φ = 0.82, p < 0.001 for meropenem; Φ = 0.70, p < 0.001 for ertapenem; Φ = 0.63, p < 0.001 for imipenem). Phylogenetic analysis revealed two distinct clusters corresponding to cfiA status, each exhibiting genetic diversity based on multi-locus sequence typing (MLST). The cfiA+ BF isolates demonstrated high-level phenotypic carbapenem resistance in the presence of upstream insertion sequences. The predominant sequence type (ST) among cfiA+ BF isolates was ST157, and 70% of ST157 isolates harbored IS1187 in the upstream region of cfiA. Gene sequencing also identified other emerging beta-lactamase genes bla_OXA-347 and bla_MUN. The 30-day all-cause mortality following B. fragilis infection was 13.3%, with independent predictors including a high Charlson Comorbidity Index (OR = 1.30; p = 0.02) and the absence of early source control (OR = 4.84; p = 0.03). This study highlights the widespread occurrence of cfiA+ BF in Hong Kong and the clinical significance of rapid cfiA detection. Continuous surveillance is essential to monitor the ongoing threat of antibiotic resistance in B. fragilis. Full article

Urban forests and green spaces are increasingly promoted as Nature-Based Solutions (NbS) to mitigate climate risks, enhance human well-being, and support resilient and sustainable cities. Focusing on the theme of optimizing urban green space ecosystem services to foster resilient and sustainable cities, this study systematically analyzes 861 relevant publications indexed in the Web of Science Core Collection from 2005 to 2025. Using bibliometric analysis and scientific knowledge mapping methods, the research examines publication characteristics, spatial distribution patterns, collaboration networks, knowledge bases, research hotspots, and thematic evolution trajectories. The results reveal a rapid upward trend in this field over the past two decades, with the gradual formation of a multidisciplinary knowledge system centered on environmental science and urban research. China, the United States, and several European countries have emerged as key nodes in global knowledge production and collaboration networks. Keyword co-occurrence and cluster analyses indicate that research themes are mainly concentrated in four clusters: (1) ecological foundations and green process orientation, (2) nature-based solutions and blue–green infrastructure configuration, (3) social needs and environmental justice, and (4) macro-level policies and the sustainable development agenda. Overall, the field has evolved from a focus on ecological processes and individual service functions toward a comprehensive transition emphasizing climate resilience, human well-being, and multi-actor governance. Based on these findings, this study constructs a knowledge ecosystem framework encompassing knowledge base, knowledge structure, research hotspots, frontier trends, and future pathways. It further identifies prospective research directions, including climate change adaptation, integrated planning of blue–green infrastructure, refined monitoring driven by remote sensing and spatial big data, and the embedding of urban green space ecosystem services into the Sustainable Development Goals and multi-level governance systems. These insights provide data support and decision-making references for deepening theoretical understanding of Urban Green Space Ecosystem Services (UGSES), improving urban green infrastructure planning, and enhancing urban resilience governance capacity. Full article

To investigate the interlayer interference mechanism during multi-layer commingled production in low-permeability gas reservoirs, this study focuses on the strongly heterogeneous reservoirs in the central Linxing area of the Ordos Basin. Laboratory-scale physical simulation experiments of commingled production were conducted on core samples from the Shiqianfeng (Q5) and Shihezi (He4) formations, along with the No. 8 + 9 coal seam. The gas production behavior, including the evolution of flow rates, the occurrence of staggered production peaks, and the resulting interlayer interference coefficients, was systematically analyzed and compared between single-layer and multi-layer commingled production scenarios. Experimental results reveal a positive correlation between cumulative gas production and layer permeability under single-layer production conditions. Specifically, the high-permeability layer (0.6470 mD) yielded 65.22 mL, whereas the low-permeability layer (0.1061 mD) produced 36.51 mL, representing a 44.02% reduction relative to the former. Under commingled production conditions, the productivity of the low-permeability layer exhibited more severe inhibition, showing declines in instantaneous production of 34.02–48.96% and cumulative production of 15.50–20.61%. These reductions substantially exceed those observed in the high-permeability layer, which ranged from 6.14% to 6.35% and from 5.00% to 8.76%, respectively. Furthermore, a greater permeability contrast results in a more pronounced difference in gas breakthrough timing. For a permeability ratio of 3, the breakthrough time difference reaches 191 s, compared to 131 s for a ratio of 2. The interlayer interference coefficient exhibits a negative correlation with the permeability contrast. When the contrast is 3, the interference coefficient for the low-permeability layer reaches 79.39%, representing an 84.51% increase relative to the coefficient observed at a contrast of 2. This indicates that larger permeability contrasts lead to more severe interference effects on low-permeability layers. These findings provide theoretical support for optimizing the efficient development of multi-layer commingled production in low-permeability unconventional gas reservoirs, highlighting the necessity of incorporating permeability contrast analysis in commingled production design. Full article

A condition-based maintenance modeling approach is proposed for a multi-state system under competing failures and imperfect repairs. The system experiences three states (normal, defective and failed) over its lifecycle. Two competing failure processes, i.e., natural degradation and external shocks, cause these state changes. If the system becomes defective, an imperfect repair is adopted to restore it to a normal state. Imperfect repairs addressing defects are mathematically characterized. Based on this, two system renewal scenarios and their occurrence probabilities are simulated and derived. The cost of downtime caused by hidden failures is then deduced. A maintenance model of the expected cost rate is constructed, and the optimal inspection period that minimizes the expected cost rate is determined. Finally, a numerical example verifies the correctness and effectiveness of the maintenance model. Full article

As a non-invasive diagnostic technique, evolved gas analysis (EGA) holds significant value in assessing the insulation conditions of critical equipment such as power cables. Current analytical methods face two major challenges: insulation materials may undergo multiple aging mechanisms simultaneously, leading to interfering characteristic gases; and traditional approaches lack the multi-label recognition capability to address concurrent fault patterns when processing mixed-gas data. These limitations hinder the accuracy and comprehensiveness of insulation condition assessment, underscoring the urgent need for intelligent analytical methods. This study proposes a deep convolutional neural network (DCNN)-based multi-label classification framework to accurately identify the gas generation characteristics of five typical power cable insulation materials—ethylene propylene diene monomer (EPDM), ethylene-vinyl acetate copolymer (EVA), silicone rubber (SR), polyamide (PA), and cross-linked polyethylene (XLPE)—under fault conditions. The method leverages concentration data of six characteristic gases (CO₂, C₂H₄, C₂H₆, CH₄, CO, and H₂), integrating modern data analysis and deep learning techniques, including logarithmic transformation, Z-score normalization, multi-scale convolution, residual connections, channel attention mechanisms, and weighted binary cross-entropy loss functions, to enable simultaneous prediction of multiple degradation states or concurrent fault pattern combinations. By constructing a gas dataset covering diverse materials and operating conditions and conducting comparative experiments to validate the proposed DCNN model’s performance, the results demonstrate that the model can effectively learn material-specific gas generation patterns and accurately identify complex label co-occurrence scenarios. This approach provides technical support for improving the accuracy of insulation condition assessment in power cable equipment. Full article

As an important factor affecting the yield and quality of grain crops and threatening grain security, traditional pest and disease monitoring can no longer meet the needs of accurate and efficient agricultural production. The development of remote sensing technology provides a new monitoring method, which is specific, accurate and efficient, and provides real-time, rapid and non-destructive spectral data information for the identification of the occurrence and severity of pests and diseases and can realize large-scale monitoring of grain crop pests and diseases. In this paper, through the statistics and analysis of the published literature on remote sensing monitoring of grain crop diseases and pests, the research hotspots and directions of remote sensing monitoring of grain crop diseases and pests are clarified. Based on this foundation, this paper systematically elaborates the mechanism underlying remote sensing-based monitoring and prediction of diseases and insect pests in grain crops. It reviews various remote sensing monitoring approaches for such diseases and pests by leveraging multi-source remote sensing data. Furthermore, it summarizes methodologies for constructing monitoring and prediction models for grain crop diseases and insect pests. Finally, the paper discusses current challenges and future development trends in this field. Full article

In semi-arid areas like Southeastern Anatolia, where agricultural productivity and water supply are extremely climate-sensitive, drought is a significant environmental and socioeconomic problem. Comprehensive assessment of drought and soil moisture dynamics is fundamental to sustainable agriculture and water security in semi-arid regions. This study analyzes drought patterns across seven provinces in the Southeastern Anatolia (GAP) region of Türkiye (Adıyaman, Diyarbakır, Gaziantep, Kilis, Mardin, Siirt, and Şanlıurfa) from 1963 to 2022, employing four drought indices (SPI, SPEI, CZI, and RDI) at multiple timescales (1-, 3-, and 12-month) to support evidence-based strategies for sustainable water and agricultural resource management. A more thorough evaluation is made possible by this multi-index and multi-scale method, which is rarely used concurrently at the provincial level. Additionally, the drought characterization was validated and enhanced through the analysis of ERA5-Land soil moisture data (1950–2022). According to the findings, the provinces with the lowest median index values and the highest frequency of extreme drought episodes are Diyarbakır and Şanlıurfa. The SPEI-12 (THW) median values showed a neutral long-term drought–wetness balance with seasonal changes, ranging from −0.0714 (Adıyaman) to 0.188 (Şanlıurfa). Particularly after 2009, soil moisture levels decreased to as low as 2–3 mm during the summer, indicating heightened evapotranspiration stress. RDI-12’s reliability in long-term drought evaluation was confirmed by its strongest correlation with other indices (r = 0.87–0.97). According to spatial research, the frequency of moderate droughts in the southwest was as high as 39%, whilst the eastern provinces experienced severe and intense droughts as high as 8%. However, with frequency above 53%, wet occurrences were more common in the east, particularly in Siirt. By clarifying long-term drought and soil moisture patterns, this study provides essential insights for sustainable irrigation planning and agricultural water allocation in the GAP region. Full article

Forest fires pose escalating threats to ecological security and public safety in Guangdong Province. This study presents a novel machine learning framework for fire occurrence prediction by synergistically integrating multi-source geospatial data. Utilizing Moderate-resolution Imaging Spectroradiometer (MODIS) active fire detections from 2014 to 2023, we quantified historical fire patterns and incorporated four categories of predisposing factors: meteorological variables, topographic attributes, vegetation characteristics, and anthropogenic activities. Spatiotemporal clustering dynamics were characterized via kernel density estimation and spatial autocorrelation analysis. An XGBoost classifier, hyperparameter-optimized through the Sparrow Search Algorithm (SSA), achieved a predictive accuracy of 90.4%, with performance evaluated through precision, recall, and $F_1$-score. Risk zoning maps generated from predicted probabilities were validated against independent fire records from 2019 to 2024. Results reveal pronounced spatial heterogeneity, with high-risk zones concentrated in northern and western mountainous areas, constituting 29% of the provincial territory. Critical driving factors include slope gradient, proximity to roads and rivers, temperature, population density, and elevation. This robust predictive framework furnishes a scientific foundation for spatially-explicit fire prevention strategies and optimized resource allocation in key high-risk jurisdictions, notably Qingyuan, Shaoguan, Zhanjiang, and Zhaoqing. Full article