Search Results (8,023)

Background/Objectives: Deep learning models for digital mammography sensor data are increasingly deployed across hospitals using different X-ray detector technologies and patient populations. Whether models trained on one sensor platform and population maintain accuracy when transferred to another has not been tested for the latest generation of mammography-specific foundation models under one controlled protocol. Methods: We fine-tuned five backbone architectures (ResNet-50, DINOv2-B14, Rad-DINO, Mammo-CLIP B5, and Mammo-FM) on CBIS-DDSM (film-digitized, USA, n = 714 validation) with three seeds, ablated a density-aware focal loss across three auxiliary weights, and evaluated transfer to three external sensor cohorts: CMMD (full-field digital, China, n = 1032), DMID (mixed digital, India, n = 509), and MIAS (film-digitized, UK, n = 322). Significance used paired DeLong z-tests with Benjamini–Hochberg FDR correction; temperature scaling tested post hoc recalibration at all transfer targets. Results: Within this single-source three-seed evaluation, ResNet-50 outperformed all four foundation models on CBIS-DDSM (AUC 0.867 vs. 0.847, 0.846, 0.813, and 0.703; all gaps p_adj < 0.05). The density-aware focal loss degraded both AUC and calibration at every weight tested. At transfer, every model lost 0.165 to 0.320 AUC points relative to in-distribution performance, with sensitivity at 95% specificity collapsing from 0.31 to 0.47 in-distribution to 0.11 to 0.22 across the three external targets. A per-seed Stouffer meta-analysis confirms that Mammo-CLIP B5 and Mammo-FM significantly outperformed ResNet-50 on DMID and Mammo-CLIP on CMMD, after BH-FDR; MIAS comparisons remained directional only. In the extremely dense subgroup (BI-RADS D4), Mammo-FM reached AUC 0.870 versus ResNet-50 at 0.842, a directional observation whose 95% CIs overlap heavily at the n = 140 sample size and which we do not interpret as a statistically supported advantage. Conclusions: In this single training-source, three-seed protocol, mammography-specific pretraining did not deliver the in-distribution AUC premium reported in the originating papers, and no architecture reached a level at which transfer deployment without local validation would be defensible. We frame these as observations specific to the present protocol rather than as broader conclusions about foundation models for mammography classification. The findings argue for sensor-stratified and population-stratified external validation and for local recalibration as practical prerequisites before clinical use. Code and weights are released under MIT license. Full article

Desert ecosystems harbor microbial communities adapted to extreme environmental conditions, including water scarcity, elevated temperatures, and intense UV radiation. Among these microorganisms, microalgae represent promising resources for agricultural applications. In this study, microalgae isolated from desert soils in Mexico were characterized by molecular (rbcL) and phylogenetic analysis, and morphological observations. They were identified as Chlorella sp. (RAD3), Nannochloris-related isolate (RAD4), and Chlorella cf. variabilis (RAD5). The effects of microalgal biomass on maize (Zea mays L.) germination and early seedling development were evaluated using a seed-priming approach. Microalgal treatments significantly improved (p < 0.05) germination-related traits, seedling vigor, shoot height, root length, and fresh and dry biomass, as compared with the control. Chlorella cf. variabilis (RAD5) was associated with reduced germination time, whereas Nannochloris-related isolate (RAD4) consistently produced the strongest responses in vigor and growth parameters. Although some variables reached their highest numerical values at 10⁸ cells/mL, similar responses were usually observed at 10⁷ cells/mL. Overall, the evaluated desert-derived microalgal preparations were associated with improved early maize seedling performance, under the evaluated conditions. Full article

Extreme climatic conditions often intensify abiotic stress factors (such as drought, salinity, heat stress, ultraviolet radiation, and soil degradation), and are increasingly limiting crop productivity and threatening global food security. Secondary metabolites (SMs), traditionally viewed as defense compounds, are now recognized as key regulators of plant adaptation to environmental stress. This review synthesizes recent advances in understanding the role of SMs as biochemical targets for improving crop resilience to climate extremes. By integrating evidence from multi-omics studies, artificial-intelligence-driven analyses, and functional genomics, we examine how stress-specific metabolic signatures and regulatory networks can be exploited for crop improvement. We further discuss the application of genome editing, synthetic biology, and metabolomics-assisted breeding to modulate the SM pathways to enhance stress tolerance. Selected case studies highlight the contribution of flavonoids, alkaloids, and terpenoids to stress adaptation in major and underutilized crops grown under salinity, drought, and low-temperature conditions. Despite significant progress, challenges remain, including metabolic trade-offs between stress tolerance and yield, regulatory constraints, and public acceptance of genetically engineered crops. By linking molecular mechanisms with applied strategies, this review provides a conceptual framework for leveraging secondary metabolism in climate-resilient agriculture and identifies key gaps to guide future research and innovation. Full article

This study established a refined, distributed SWAT modeling framework that integrates elevation-band and snowmelt modules to reconstruct the alpine hydrological and sediment cycles of the Koshi River Basin (KRB) over the period 1990–2024, with climate scenarios constructed using the delta change approach. The KRB, a major transboundary watershed traversing China, Nepal, and India, was selected owing to its critical hydro-climatic role under the destabilizing “Asian Water Tower”; it generates substantial sediment yield, hosts the densest concentration of hydropower potential within the Ganges system, and spans an extreme vertical gradient from Mount Everest to the southern alluvial plains. Results reveal accelerated warming at a rate of 0.21 °C per decade and an overall warming–wetting trend, punctuated by an abrupt interdecadal shift around 2015. Precipitation dominated interannual streamflow variability, with enhanced rainfall triggering basin-wide sediment surges that overwhelmed the natural buffering capacity of the land surface. Conversely, rising temperatures intensified actual evapotranspiration, markedly depleting soil water and reducing total water yield and monsoon runoff, although sustained snow and glacier melt effectively elevated the dry-season low-flow baseline. The integrated climate forcing reshaped the disparity between hydrological extremes, imposing severe seasonal eco-hydrological stress that manifested as a pre-monsoon deficit in terrestrial green water and acute summer sediment outbursts for aquatic habitats. Furthermore, the flood regime exhibited an altered distribution, with mid-to-high frequency floods enhanced while low-frequency extreme flood peaks declined. The hydro-sedimentological regime consequently exhibits pronounced nonlinear responses to climate change, providing a critical, threshold-based scientific foundation for adaptive transboundary water resource management. Full article

The rapid transition toward renewable energy requires transparent and spatially explicit methods for identifying suitable photovoltaic (PV) development areas. This study develops a geospatial artificial intelligence (GeoAI) decision-support framework for PV site suitability assessment in Rajasthan, India. Eleven harmonized predictors were used: global horizontal irradiance (GHI), photovoltaic power output (PVOUT), temperature, wind speed, aerosol optical depth (AOD), elevation, slope, albedo, land use/land cover (LULC), distance to roads, and distance to power lines. Reference labels were generated from an explicit rule-derived suitability index, class thresholds, and exclusion logic; therefore, the machine-learning task was to reproduce a transparent suitability framework rather than to predict observed PV yield or project-level performance. Extreme Gradient Boosting (XGBoost) was compared with simpler baseline models, evaluated using random and spatial-block validation, and interpreted using SHapley Additive exPlanations (SHAP). Independent overlays with known solar-installation records, presence-background robustness testing, and uncertainty/sensitivity analysis were used to examine spatial plausibility, spatial autocorrelation, deterministic label effects, and parameter uncertainty. The resulting outputs include pixel-level suitability zones, contiguous candidate polygons, district-level capacity-oriented summaries, and planning-priority classes. The framework is intended as a risk-aware regional screening tool: high model agreement indicates consistency with the constructed suitability labels, while final project decisions require parcel-scale land, grid, environmental, social, and economic assessment. Full article

The Jabal Ali Marine Sanctuary, Dubai, is one of the most important marine protected areas (MPAs) in the UAE. The Arabian Gulf is characterised by extreme environmental conditions, including high temperatures and hypersaline waters. These conditions, combined with increasing anthropogenic pressures from coastal development projects such as desalination plants, energy plants and the Palm Jebel Ali development, may influence the pelagic ecosystems of MPAs. This study examined seasonal variability in phytoplankton communities and environmental conditions between summer (June 2017) and winter (December 2017), with particular emphasis on the interactions between temperature-driven stratification, hypersaline conditions, and phytoplankton community structure, abundance, and diversity. The AZTI (AZTI Tecnalia Marine Research Centre) Marine Biotic Index indicated predominantly “Good” to “High” ecological status of the pelagic ecosystem, indicating favourable environmental conditions. Potentially harmful algal bloom taxa, including Pseudo-nitzschia and Dinophysis, were detected at low abundances. Summer surveys recorded higher total species richness (44 vs. 34 species) and greater phytoplankton abundance (mean 68.6 vs. 49.8 cells/L) compared to those in winter. Diatoms dominated the assemblages in both seasons, accounting for 62–69% of the recorded species, while distinct spatial zonation patterns reflected habitat heterogeneity. The observed seasonal and spatial variability highlight the importance of incorporating temporal and spatial dimensions into management strategies. As the first pelagic phytoplankton assessment conducted in an MPA, this study provides important baseline data for understanding phytoplankton ecology in one of the world’s most environmentally extreme marine ecosystems. The findings contribute to evidence-based management under increasing climate change and anthropogenic pressures. However, because sampling was limited to the two principal climatic seasons, the study characterises inter-seasonal variability rather than a complete annual succession cycle. Additional surveys during spring and autumn are recommended to fully resolve seasonal succession dynamics. Overall, the findings support the continued protection of the sanctuary as an important biodiversity reservoir and a potential reference site for assessing marine ecosystem responses to environmental conditions. These findings are directly relevant to the environmental sustainability agenda of the Dubai 2040 Urban Master Plan, which prioritises the protection and expansion of the emirate’s nature reserves and the safeguarding of marine and coastal biodiversity. By establishing the first pelagic phytoplankton baseline for the sanctuary, this study provides an evidence base for monitoring and managing marine protected areas in line with this long-term framework. Full article

With ongoing urban warming and the increasing frequency of extreme heat events, thermal comfort at highly exposed public spaces like bus stops has attracted significant attention. However, existing studies largely rely on linear assumptions and limited environmental variables, leaving the complex, multidimensional mechanisms driving thermal perception unclear. This study investigates the nonlinear impacts of microclimate, urban morphology, and station design on passengers’ thermal perception during summer in Chongqing, China. Drawing on field measurements, questionnaire surveys, and spatial data, linear regression was first applied to estimate neutral temperatures and acceptable thermal ranges. Subsequently, an interpretable machine learning framework integrating XGBoost and SHAP analysis was developed to explore the nonlinear effects and interactions mechanisms among these variables. The results reveal a dual regulatory pattern. Mechanism variables exhibit distinct nonlinear thresholds, with wind speeds above 0.98 m/s showing cooling associations and PET values exceeding 34.70 °C corresponding to more rapid increases in thermal discomfort. Concurrently, urban morphology and station design factors contextually modify these direct effects by altering their magnitude and direction. Furthermore, significant spatial heterogeneity in thermal adaptation was observed, with neutral temperatures ranging from 23.19 to 31.01 °C. These findings provide a basis for developing adaptive and context-specific thermal environment management strategies for urban bus stops. Full article

This study systematically investigates the hot deformation behavior of TC18 alloy under the conditions of deformation temperatures of 720–840 °C and strain rates of 0.001–1 s⁻¹. Based on the stress–strain data obtained under the aforementioned process parameters, a support vector regression (SVR) model was established and further optimized by using a Stacking algorithm to enhance predictive accuracy. Although SVR and Stacking techniques have been applied previously in material constitutive modeling, this paper presents a systematic optimization framework specifically for TC18, integrating comprehensive experimental data, kernel selection, hyperparameter tuning, and Stacking-based model fusion. The polynomial kernel function was identified as optimal, and hyperparameters were tuned via grid search combined with five-fold cross-validation, which is determined as {C = 1000, coef0 = 1, d = 5, ε = 1, γ = 1}. The Stacking-SVR model exhibits significantly improved fitting and generalization performance compared to Poly-SVR, Arrhenius, XGBoost and MLP, with RMSE, MAPE, and R² metrics of 2.7882, 0.0110, and 0.9973 on the training set, and 2.7956, 0.0169, and 0.9982 on the test set, respectively. Additionally, the proportion of samples with relative errors within 5% reaches 98.7% for the training set and 94.83% for the test set. These results indicate that the proposed framework not only possesses extremely high predictive accuracy, but also ensures strong generalization ability and interpretability in practical applications. Full article

Introduction: The fartet (Apricaphanius iberus) and the samaruc (Valencia hispanica) are two endemic fish species from the Valencian Community that have experienced significant population declines due to habitat degradation, competition with invasive species, and the impacts of climate change. Despite their critical conservation status, key aspects of their population dynamics and reproductive biology remain poorly understood. Objective: This study aimed to assess the resilience of their embryos to water stress through diapause-like mechanisms. Methodology: For studying the embryonic arrestment, eggs were collected from captive populations and subjected to different incubation periods (1, 3, 7, 10 and 14 days) on different substrates (commercial sand and filter paper). Hatching rates were analyzed in relation to the duration of exposure to stress water conditions and the type of substrate used. Results: The experiments conducted demonstrated that the embryos of both species were able to withstand water stress conditions (eggs out of the water). In the case of the samaruc, the results showed that eggs collected in both May and June could resist water-stress conditions for at least 10 days, exhibiting hatching rates of 100% during this period, which decreased to 50% by day 14. Regarding the fartet, embryos from eggs collected in May were able to survive up to 3 days under water-stress conditions, with hatching rates of 100%. In contrast, embryos from eggs collected in June showed greater resilience to water stress, with high hatching rates of 60–100% at days 7 and 10. Conclusions: These results suggest that, although a mechanism like embryonic diapause may be present in these species, its effectiveness as an adaptive strategy may depend on multiple environmental factors not controlled in this study, such as temperature, oxygen availability, and water salinity. The absence of hatching after prolonged incubation periods indicates that, if a diapause mechanism exists in these species, it may not be as efficient as in other annual cyprinodontiforms adapted to extremely fluctuating environments. These results highlight the importance of adaptive management measures to mitigate the effects of climate change and ensure the long-term persistence of both species. Full article

In recent decades, summer extreme high-temperature (EHT) events in the Sichuan–Chongqing (SC) region of southwestern China have become increasingly frequent under global warming. Carbon dioxide removal (CDR) is considered a key strategy for achieving the temperature targets of the Paris Agreement; however, the response of regional EHT events to CDR remains poorly understood. Based on CN05.1 observations and idealized CO₂ ramp-up and ramp-down experiments from the CMIP6 Carbon Dioxide Removal Model Intercomparison Project (CDRMIP), this study investigates the historical characteristics of summer EHT events over eastern SC and their responses to CDR. The results show that historical EHT events have become more frequent, longer-lasting, and more intense, indicating an overall intensification of regional high-temperature risk. Under idealized CO₂ pathways, regional mean temperature and EHT frequency exhibit pronounced asymmetric and hysteretic responses, with positive anomalies persisting even after CO₂ returns to its initial level. This asymmetric response is closely associated with the enhanced slow oceanic response during the ramp-down period. Stronger El Niño-like and Indian Ocean Dipole-like SST warming intensifies the South Asian High and western Pacific subtropical high, favoring elevated summer temperatures and increased EHT events over eastern SC. Soil moisture also heats the atmosphere by altering the surface latent heat flux in the southwestern part of the study region during ramp-down period. These findings not only improve the understanding of regional extreme event responses in the SC region under carbon neutrality, but also confirm the positive effect of carbon neutrality targets on mitigating regional extreme climate change, thereby highlighting the urgent need to control CO₂ emissions. Full article

Journal bearings are prone to failure due to lubrication state degradation under extreme operating conditions. To address the unclear transition mechanism and undefined state boundaries under insufficient lubrication, a coupled tribological model of engine journal bearings was established. Through parameter analysis and dynamic failure mechanism study, the effects of radial clearance, temperature, rotational speed, load, and surface roughness on the lubrication state transition were revealed. The results indicate that radial clearance, oil temperature, rotational speed, applied load and surface roughness are all decisive factors for lubrication transition, and every parameter has its unique critical threshold; once exceeding the limit, the oil film integrity is damaged and the lubrication rapidly shifts from mixed lubrication toward boundary lubrication. After crossing critical thresholds, aggravated asperity contact further triggers continuous temperature rise and viscosity reduction, which may induce closed-loop thermal deterioration and eventually accelerate bearing failure. The research findings provide a theoretical basis for robust design and operational safety monitoring of journal bearings. Full article

Introduction: Coastal fishes can adapt to water warming and hypoxia; however, acute tolerance does not necessarily predict longer-term performance and survival. This may be especially important in sedentary, site-faithful species with limited escape to escape increasingly unfavorable habitats. We assessed the climate-related stress responses of the Lusitanian toadfish, Halobatrachus didactylus, a benthic estuarine fish from the Northeast Atlantic, to water warming and hypoxia. Objectives: We aimed to determine the aerobic energy budget, thermal limits (CTmax), and critical oxygen tension (Pcrit), as well as blood indicators of metabolism, altered physiology and systemic stress, as proxies for whole-organism homeostatic state, thereby informing future ecophysiological assessments and bioindicator development in a context of environmental change. Methodology: We determined standard, routine, and maximum metabolic rates; aerobic scope; and critical thermal maximum (CTmax) and critical oxygen (Pcrit) thresholds on a set of 134 individuals ranging from 12 to 160 g in weight. On a different set of individuals (n = 48; 76.3 ± 2.6 g; 16.1 ± 0.18 cm), we simulated 30 days of seasonal scenarios combining low and high temperature with normoxia or hypoxia, followed by integrated metabolic, hematological, biochemical, and multivariate analyses. Results: Acute trials showed high short-term resilience: H. didactylus had an exceptionally low standard metabolic rate and routine metabolic rate, high CTmax (34.82 ± 0.66 °C), and strong hypoxia tolerance (Pcrit 0.59–1.97 mg O₂ L⁻¹), although smaller individuals were more sensitive. After 30 days, however, warming more than doubled standard and routine metabolic rates, while warm hypoxia reduced metabolic output relative to warm normoxia, consistent with metabolic depression under compounded stressors. This treatment also showed shifts in glucose, liver mass, red blood cell count, and hematocrit, identifying warm, oxygen-poor water as the most physiologically costly scenario for this species. Conclusions: Together, these results show that high acute tolerance does not guarantee resilience to climate change. In sedentary fishes, survival may depend less on surviving extremes than on maintaining energetic balance, oxygen transport capacity, and physiological homeostasis in increasingly warm, oxygen-poor coastal habitats. Full article

Introduction: Climate warming and drought are intensifying thermal stress in Mediterranean freshwater systems, potentially favoring invasive fish with broad physiological tolerance. Extended environmental tolerance and increased aerobic scope are indicative of the potential to sustain, perform and disseminate in challenging conditions. Objective: We aimed to determine the thermal scope of the invasive Australoheros facetus inhabiting southern Portuguese drainages using an array of physiological proxies. Methodology: We evaluated the thermal biology of the species across a wide temperature gradient to test how warming affects metabolic performance, thermal tolerance, and biochemical status. Fish collected from Algarve watercourses were exposed to 5, 10, 15, 20, 25 and 35 °C (n = 15 per condition, 10–60 g) for at least a week, and intermittent respirometry was used to determine standard metabolic rate (SMR), maximum metabolic rate (MMR) and aerobic scope (AS). Group Q10 was derived from metabolic rates. Plasma and tissue biomarkers of energy metabolism and oxidative stress were analyzed. Critical thermal maximum (CTmax) was assessed in fish acclimated for a week at 10, 20 and 30 °C (n = 10) using a 1 °C/min thermal ramp. Results: Intermediate temperatures (15–25 °C) supported the best overall physiological performance, combining stronger aerobic capacity with higher antioxidant protection. In contrast, 30–35 °C imposed clear physiological costs: maintenance metabolism increased disproportionately, aerobic scope declined, and cellular protection weakened, indicating the onset of heat stress. Despite this, A. facetus showed marked thermal plasticity, with CTmax increasing significantly with acclimation temperature. Fish acclimated to 30 °C had higher CTmax than fish acclimated to 20 °C and 10 °C, although the thermal safety margin decreased progressively as the acclimation temperature rose. Liver antioxidant activity also peaked at intermediate temperatures and declined at the warmest treatments, reinforcing the mismatch between acute tolerance and sustained performance. Conclusions: These results show that A. facetus is highly heat tolerant but that tolerance comes with energetic and cellular trade-offs near upper thermal limits. Despite this limitation at extreme conditions, the combination of broad tolerance and functional performance under warm intermediate conditions may help to explain its invasion success and stand as a competitive advantage in increasingly hot low-flow Iberian freshwater ecosystems. Full article

Frequent seasonal phase transitions in cold and arid lakes require different remote sensing indices for frozen and open-water periods, complicating the use of traditional empirical indices for automated monitoring. To address this challenge, this study proposes an intelligent indexing framework integrating the heuristic reasoning of Large Language Models (LLMs) with Random Forest (RF) feature selection. Leveraging the Google Earth Engine (GEE) and Landsat 8 data from Ulansuhai Lake, five LLMs such as Gemini and ERNIE were employed to generate candidate spectral indices based on typical sample spectra. Optimal band combinations were identified via RF importance, and Land Surface Temperature (LST) was incorporated as a physical constraint for unified cross-seasonal classification and determine the optimal threshold. Results show that the LLM-derived ERNIE-WISI and Gemini-WISI exhibit high robustness. During the freezing period, ERNIE-WISI significantly outperformed other indices, achieving an Overall Accuracy (OA) of 89% and a Kappa of 0.86. Spatially, it yielded snow and ice mapping with clear textures and low commission errors. During the non-freezing period, ERNIE-WISI achieved an OA of 95% with a Kappa of 0.84. While Gemini-WISI achieved an OA of 94% with a Kappa of 0.80, performing comparably to MNDWI. Notably, ERNIE-WISI effectively suppressed background interference in complex landscapes like narrow channels and aquaculture areas, maintaining high geometric fidelity and spatial continuity. A key advantage of ERNIE-WISI is its consistent performance without seasonal threshold adjustments. Aligned with the AI for Science paradigm, this methodology bridges AI-driven heuristic discovery and physical remote sensing, offering a robust, transferable solution for long-term dynamic lake monitoring in extreme environments, thereby facilitating sustainable water management. Full article

To achieve scientific cost forecasting, this study investigates structural changes in mining cost driven by the widespread adoption of mechanized mining, increased mining depths, and significant operational variations. Based on the backpropagation (BP) neural network, this study systematically analyzes the cost-composition characteristics of modern mining operations and applies activity-based costing to achieve refined cost accounting for each mining operation unit. Ten key influencing factors, including working space, stope temperature, stope depth, haulage distance, worker seniority and work efficiency, scraper efficiency, equipment service life, fuel and lubricant consumption rates, are identified by analyzing cost variation patterns. Principal component analysis (PCA) is used to reduce the dimensionality of the ten factors to simplify this model and enhance prediction accuracy. The PCA-BP neural network mining cost forecasting model is built with the principal components extracted as input variables. Actual cost data from an underground metal mine in Shandong Province is used for our model training and validation, with adopting linear regression, eXtreme Gradient Boosting (XGBoost), and a traditional BP neural network as the comparison models for performance evaluation. Our prediction results indicate that the PCA-BP model achieves an average relative error of 3.80% and a root mean square error of 1.43, both significantly outperforming the comparison models. The results demonstrate superior predictive accuracy and stability of our model. Validated with data from a typical deep mechanized gold mine in eastern China, the PCA-BP cost forecasting model requires parameter retraining based on local production conditions for applications in other regions. This study confirms that the model aligns well with the cost characteristics of modern underground metal mines and produces effective predictions, offering reliable quantitative support for the development of cost control strategies and optimization of cost planning in mining enterprises. Full article