Search Results (1,282)

With the increasing integration of grid-friendly concentrated solar power (CSP) plants into high-proportion new energy power systems, the system is confronted with challenges such as insufficient regulation capability and power balance difficulties. To address these issues, this paper proposes a multi-stage optimal regulation strategy for CSP–wind power systems based on adaptive step-size model predictive control (ASMPC), from the perspectives of tapping transmission line current-carrying capacity and coordinating system regulation resources. This strategy first establishes an electro–thermal–mechanical coupling dynamic line rating (DLR) model to characterize line safety margins, then constructs an optimization decision-making model aiming at minimizing the total multi-stage coordinated scheduling cost and adopts ASMPC to dynamically adjust the control step size, effectively improving scheduling accuracy and real-time correction capability. Simulation results based on the modified IEEE 39-bus system show that the proposed method reduces the total system cost by 26.8% (nearly 30%), increases the CSP unit output ratio by 27.9%, and decreases the average grid load rate by 12.6 percentage points. The proposed strategy can effectively mitigate the impact of source-load uncertain fluctuations and significantly improve the economic operation level of the CSP–wind power combined system. Full article

Integrated care is increasingly shaped by digital infrastructures, data governance, and AI-enabled analytics, yet the relevant literature remains fragmented across health-services research, digital health, and machine learning. This article reports a scoping review, conducted in line with PRISMA-ScR guidance, that maps how integrated care models have evolved conceptually, what digital and AI-enabled infrastructures support them, how their clinical, economic, and equity impacts can be evaluated, and what current implementations imply for sustainable scaling. We searched PubMed, Scopus, Semantic Scholar, and Crossref (retrieval date 31 October 2025; forward screening to 31 March 2026) and added grey literature from named policy bodies. The searches identified 15,189 records, reducing to 11,789 after intra- and cross-source deduplication and grey-literature integration; 620 full texts were assessed and 192 were included in the synthesis. Four domains were synthesised: conceptual foundations of integrated care, AI and multimodal analytics, implementation barriers, and digital-governance foundations. We chart the field using a Type I–V maturity scheme (disease, cohort, whole-system, digital-integrated, learning), benchmarked against the Rainbow, MacColl, EMRAM/AMAM, and NHS ICS models. Most deployments cluster at digitally integrated but only weakly adaptive Type IV; recurrent failure modes—temporal blind spots, maintenance debt, semantic drift, and governance gaps—block progression to Type V, and high-profile clinical-AI failures illustrate the cost of attempting Type V analytics on Type IV-or-worse infrastructure. A walk through nine world regions maps each to its current Type I–V position and shows that organisational and payment integration—not digital sophistication alone—is currently the dominant driver of progress. The COMFORTage Integrated Care Model Library is positioned as a workflow of AI agents orchestrating predictive, preventive, and personalised care across the integrated-care lifecycle rather than as a single federated-learning programme. The review positions AI-enabled integrated care less as a finished model than as an emerging design space requiring longitudinal data assets, stewarded model lifecycles, accountable governance, and outcome-based contracting for clinically useful, equitable, and trustworthy learning systems. Full article

Category-level generalization for robotic garment manipulation, such as bimanual smoothing, remains a significant hurdle due to high dimensionality, complex dynamics, and intra-category variations. Current approaches often struggle, either overfitting with concurrently learned visual features for a specific instance or, despite Category-level perceptual generalization, failing to predict the value of synergistic bimanual actions. We propose the Feature-Conditioned Bimanual Value Network (FCBV-Net), operating on 3D point clouds to specifically enhance intra-category policy generalization—generalizing across unseen variations within a single topological class, as distinct from cross-category transfer—for garment smoothing. FCBV-Net conditions bimanual action value prediction on pre-trained, frozen dense geometric features, ensuring robustness to intra-category garment variations. Trainable downstream components then learn a task-specific policy using these static features. In simulated PyFlex environments using the CLOTH3D dataset, FCBV-Net demonstrated superior intra-category generalization. It exhibited only an 11.5% efficiency drop (Steps80) on unseen garments compared to 96.2% for a 2D image-based baseline, and achieved 89% final coverage, outperforming an 83% coverage from a 3D correspondence-based baseline that uses identical per-point geometric features but a fixed primitive. These results highlight that the decoupling of geometric understanding from bimanual action value learning enables better intra-category generalization. Full article

Objectives: Vancomycin is crucial for treating severe Gram-positive infections, but its narrow therapeutic index complicates dosing. Trough monitoring may inadequately reflect exposure, while AUC-guided dosing, although recommended, is often impractical. Alternative metrics such as the time in therapeutic range (TIR) and volatility index may reflect exposure dynamics. Augmented renal clearance (ARC) further challenges vancomycin therapy in Intensive Care Unit (ICU) settings. This study evaluated trough-based exposure metrics and their associations with ICU mortality and acute kidney injury (AKI). Methods: We retrospectively analyzed 109 ICU patients with sepsis receiving vancomycin. Exposure was assessed using mean trough concentrations, TIR (proportion of troughs within predefined ranges), and the volatility index, defined as the intra-individual standard deviation divided by the mean trough concentration (SD/mean). Outcomes were ICU mortality and AKI. Associations were evaluated using multivariable regression, bootstrap resampling, and restricted cubic splines. Results: TIR _>15 was independently associated with higher mortality (adjusted OR 3.88; p = 0.0326) and AKI stage II–III (adjusted OR 5.63; p = 0.0068). Higher mean troughs correlated with AKI stage II–III, whereas higher volatility showed an inverse association (adjusted OR 0.15; p = 0.0240). ARC (4.6%) occurred exclusively in younger patients and predicted subtherapeutic exposure (TIR _<10, p = 0.0485). Conclusions: Sustained troughs >15 mg/L were associated with mortality and nephrotoxicity, while the most favorable outcomes were descriptively observed at mean trough levels of approximately 8–12 mg/L, suggesting a possible narrow exposure range that requires prospective validation. These findings highlight the limitations of trough-based monitoring alone; the trough-derived metrics should be regarded as exploratory rather than validated decision-making tools. Full article

Accurate vineyard yield estimation is essential for harvest planning, resource allocation, and economic decision-making, particularly under conditions of high spatial variability. Traditional sampling-based methods are labor-intensive, destructive, and prone to error, especially in high-productivity free-canopy systems. This study developed and evaluated predictive models for commercial irrigated vineyards of Carménère and Chardonnay in Chile’s Maule Region across two growing seasons (2023–2025). Structural yield components, physiological measurements, and UAV-derived multispectral indices (NDVI, GNDVI, NDRE) were collected from georeferenced sampling grids. Modeling approaches included linear regression, stepwise selection, and machine learning algorithms (Random Forest, Multilayer Perceptron). Validation results showed that cluster number was the primary driver of yield variability, explaining up to 40% of variation. Incorporating physiological and spectral variables improved accuracy, with the best models (least squares and MLP) achieving R² values up to 0.66 and reducing errors to 12–15%. Spatial yield maps reproduced intra-vineyard variability patterns, demonstrating that integrating plant-level and canopy-level data substantially enhances yield prediction. These findings provide a robust framework for precision viticulture applications. Full article

Clinical evaluation of scoliosis primarily relies on the Cobb angle measured on standing whole-spine radiographs. However, manual measurement is affected by intra- and inter-observer variability caused by differences in end-vertebra selection, endplate definition, and vertebral boundary interpretation. In addition, low radiographic contrast and anatomical overlap can hinder accurate identification of the spinal contour. In clinical screening, rapid three-level severity classification with reduced false negatives serves as a complementary function to precise quantitative measurement, supporting case triage and missed-detection prevention. This study proposes a boundary-enhanced YOLO-based instance segmentation framework—where ‘boundary-enhanced’ refers to the reinforcement of spinal contour boundary representation through the DeepLabV3+-based segmentation head—for three-level scoliosis severity screening using clinician-assigned severity labels derived from Cobb angle measurements. Unlike semantic segmentation, which may cause class fragmentation within a single spine, the proposed method defines the entire spine as one anatomical instance and predicts a single severity label based on the global contour structure. Class-balanced offline augmentation, background-only negative samples, attention modules, and segmentation heads were comparatively evaluated. Results showed that background-only negative samples reduced false negatives, and CBAM improved accuracy while maintaining a practical model size and near-real-time inference speed under the tested environment. DeepLabV3+ provided the most stable contour reconstruction. The final model improved both contour extraction and three-level severity screening performance, suggesting that the proposed framework may be potentially useful for assisting scoliosis screening. However, further external validation and prospective evaluation are required before clinical deployment. Full article

Satellite-derived bathymetry (SDB) provides a cost-effective means for mapping shallow-water depths, yet its scalability and cross-regional generalizability remain challenging in optically complex coastal environments. This study systematically evaluates machine learning (ML) and deep learning (DL) approaches for transferable SDB over the 0–20 m depth range using multispectral Sentinel-2 imagery. A Random Forest model and four deep learning architectures–ResNet-50, ResNet-101, EfficientNet-B4, and ConvNeXt-Large–are developed and trained using data from Pratas Island (South China Sea) and selected reef regions of the Great Barrier Reef (GBR), and subsequently evaluated on spatially independent intra-regional and cross-regional test areas to assess generalization performance. Model sensitivity is investigated with respect to key training configurations, including loss-function design and data-splitting strategy. To enhance shallow-water learning, we introduce a Smooth Weight Function (SWF)-weighted RMSE loss that emphasizes near-surface depths and compare it with conventional RMSE and relative percentage error (RPE) objectives. In terms of training data, preserving spatial continuity during training substantially improves both numerical accuracy and structural consistency of predictions compared with random patch splitting. While the Random Forest model performs competitively in intra-regional tests, its accuracy degrades under cross-regional transfer (RMSE increasing from 1.53 m to 2.99–3.78 m). Deep learning models, although not always outperforming Random Forest in intra-regional settings, exhibit greater robustness to geographic shift. Using the spatially continuous training strategy, intra-regional RMSE ranges from 1.15 to 1.92 m over the full 0–20 m range, with shallow-water RMSE as low as 0.26 m for depths ≤ 3 m. Cross-regional transfer to geographically independent reefs yields moderate RMSE values of approximately 2.46–2.98 m (0–20 m range), indicating that geographic transfer remains challenging despite meaningful improvements over Random Forest. We further benchmark the proposed architectures against a task-specific bathymetry network using the public MagicBathyNet dataset. Under a unified 0–16 m shallow-water configuration using aerial RGB imagery, the proposed models achieve RMSE values between 0.19 and 0.22 m, outperforming both the baseline U-Net and the transformer-based bathymetry architecture while using substantially fewer parameters. In addition, we exploit multi-temporal repeat imagery for both training and inference, which increases training diversity and improves robustness to temporal variability arising from changing sun angles, atmospheric conditions, water properties, and tides. During inference, predictions from multiple repeat images are aggregated using the median to reduce noise and improve stability. Finally, we release optimized network architectures and pretrained weights to facilitate scalable application to new sites. This work demonstrates a practical pathway toward transferable, large-area SDB from multispectral satellite imagery using deep learning. Full article

Rapid urbanization and associated greenhouse gas emissions pose severe challenges to global climate goals. Accurately estimating urban carbon emissions at fine administrative scales is a critical prerequisite for spatially differentiated mitigation policies and achieving carbon neutrality. However, while current research has validated the feasibility of using nighttime light (NTL) remote sensing for carbon estimation, most studies predominantly focus on macro scales, paying limited attention to intra-urban spatial heterogeneity and the value of high-resolution imagery. Using Nanjing, China, as a case study, this study examines the optimal scale, model, and data source for estimating urban total carbon emissions. NTL features from NPP/VIIRS and Luojia1-01 imagery were extracted at the district and township levels. Spatial lag and spatial error models were compared, and geographically weighted regression was further applied at the township level. The results show that urban carbon emissions in Nanjing exhibit clear scale effects and spatial non-stationarity. At the township level, the total indicator (TCE-TNLI) better reflects emission expansion in peripheral areas, while the intensity indicator (CI-ANLI) shows better predictive performance and robustness. With high-resolution Luojia1-01 imagery, the intensity model further reduces the effects of pixel saturation and administrative scale differences, achieving better model performance. These findings establish a robust methodological framework for fine-scale urban carbon accounting, demonstrating that integrating high-resolution imagery with intensity-based models is crucial for supporting spatially differentiated low-carbon planning in high-density megacities. Full article

Early identification of at-risk students is crucial for timely pedagogical intervention. Determining which assessments instructors should prioritize is complicated by the fact that different eXplainable-AI (XAI) methods can produce conflicting rankings for the same predictive model. We develop a framework combining a sequential GRU model with two complementary XAI techniques, Gradient SHAP (attribution) and DiCE (counterfactuals), and evaluate it in a foundational Data Structures and Algorithms course. The framework produces predictions and explanations for every prefix length throughout the semester and quantifies inter-method agreement and intra-method stability using three disagreement metrics. Intersecting the top-k features identified by both methods isolates a compact subset of assessments whose predictive role is confirmed across two fundamentally different explanation mechanisms. We interpret this cross-method agreement as a heuristic that increases confidence in identified features relative to single-method results, though not as evidence of causal validity. For individual students, the framework uses the intersection of the two types of explanations when it is non-empty; otherwise, the instructor chooses between SHAP’s diagnostic view and DiCE’s prescriptive view, with an optional check against the top-k list. The resulting guidance is less susceptible to method-specific biases than analyses relying on a single method. Full article

While promising results have been demonstrated for deep learning-based breast cancer diagnosis using mammography, problems persist in approaches that rely primarily on visual information. These problems include inadequate performance across diverse clinical centers, various imaging protocols, scanner types, and patient distributions. Here, we introduce Cross-Center Vision–Language Transformer (CC-VLT), a framework that integrates mammograms and clinical text to enable more robust, guided diagnosis. The framework incorporates a vision transformer for mammograms, a text transformer for salient clinical descriptors, bi-directional cross-modal attention for semantics, and a cross-center feature regularization approach to address the challenge of inter-institutional domain shifts. The framework is tested on a leave-one-center-out basis across several public mammography datasets and significantly outperforms strong baseline models in both intra- and cross-center evaluations. Our framework achieved an accuracy of 90.7% with an intra-center ROC–AUC of 0.951 and cross-center ROC–AUC results of 0.912, 0.927, and 0.934 on the CBIS-DDSM, INbreast, and VinDr-Mammo datasets, respectively. Reliability of the malignancy probability predictions improved, as evidenced by a diminished Expected Calibration Error and Brier Score. Our framework, by designing an effective integrated vision–language interaction model and implementing a cross-center feature regularization approach, sets a benchmark for robust breast cancer diagnosis across diverse clinical environments. Full article

The “experimenter gender effect” is a pervasive confound in rodent behavioral neuroscience: the sex of the human handler alters stress, social, and pharmacological responses via olfactory cues and conserved neural circuits. Whether this effect extends to dog—a species co-domesticated with humans for over 15,000 years—has never been systematically tested. Here, we examined sex-biased social preferences in Beagle dogs during both intra- and cross-species interactions, and asked whether baseline neuroendocrine states predict such preferences. Thirty-four adult Beagles (17 males, 17 females) from a standardized laboratory colony underwent social interaction tests with same and opposite-sex conspecifics and with male and female experimenters. Baseline plasma corticosterone, serotonin (5-HT), and dopamine were measured by ELISA. Results indicated that Beagles did not exhibit significant sex-based preferences for either gender of conspecifics or human experimenters in either dog–dog or human–dog social interaction tests (all p > 0.05). However, males showed markedly higher baseline corticosterone, 5-HT, and dopamine than females (all p < 0.0001), a hormonal dimorphism that did not correlate with any behavioral measure in Spearman correlation analysis (p > 0.05). Nevertheless, this study has several limitations: only baseline hormone levels were measured (not stress-induced responses), behavioral tests involved only low-stakes affiliative interactions, and only one breed was studied under standardized conditions. These results suggest that Beagle dogs may lack experimenter-gender preference in social interactions, exhibiting stable, gender-neutral social behavior despite profound underlying hormonal differences. This decoupling of internal state from behavioral output suggests that domestication may have shaped a social phenotype resistant to the experimenter gender effect, supporting the Beagle as a valuable translational model with a stable baseline and low susceptibility to confounding social cues, making it suitable for research on affective and social-cognitive disorders. Full article

Behavioural analytics remain a dominant approach for modelling learner engagement and predicting performance in digital learning environments. However, existing approaches are largely retrospective, relying on observable behavioural outcomes rather than modelling the underlying cognitive state dynamics that evolve during sustained learning. This study proposes a nonlinear state-space modelling framework that formalises the interaction between cognitive fatigue, attention, and learning as a continuous-time dynamical system. Fatigue is modelled as a latent state governed by load–recovery dynamics, attention is represented as a fatigue-coupled cognitive resource, and learning accumulation is expressed as an attention-mediated process under saturation constraints. The model is discretised and empirically estimated using time-indexed webcam-derived pilot data (N = 63) and further validated using a large-scale intervention dataset (N = 1245). Parameter estimation is performed using regression-based approximation of the discretised state equations, with cluster-robust inference applied to account for intra-session dependencies. The webcam-derived features were pre-processed using temporal windowing and normalisation to ensure consistency across sessions. The swarm-optimised intervention was implemented through adaptive control of instructional load and recovery scheduling, enabling real-time regulation of fatigue progression. Empirical results demonstrate statistically significant model validity, with fatigue dynamics showing moderate explanatory capability (R² = 0.543, p < 0.001) and attention dynamics also significant (R² = 0.499, p = 0.004). At the system level, adaptive intervention significantly reduced fatigue and improved learning performance (t(1244) = 14.34, p < 0.001). The findings suggest a transition from retrospective behavioural modelling toward anticipatory cognitivestate regulation, contributing toward a computational foundation for fatigue-aware adaptive learning systems. Full article

Automated valet parking requires reliable long-range slot searching and precise low-speed docking in confined structured lots. This paper proposes a hierarchical cruise-to-park framework that combines nonlinear model predictive control (NMPC) for predefined-route cruising with a Twin Delayed Deep Deterministic Policy Gradient (TD3) agent for terminal parking. The system is implemented in a structured Simulink environment with Unreal Engine-based geometry-aware sensing modules. During cruising, a camera-based module detects available slots and triggers the transition to parking. The NMPC uses a custom cost function to improve tracking on curved approaches, while the TD3 policy uses LiDAR feedback and reward shaping with an explicit time penalty to encourage efficient, stable docking. Simulation results demonstrate smooth phase transition, accurate cruising, and effective terminal parking in the training slot. Validation on six previously unseen target slots within the same parking-lot environment shows encouraging intra-lot target-slot transferability without retraining. Additional PPO and SAC comparisons and a time-penalty ablation further evaluate the relative learning performance and the effect of reward design, supporting the proposed architecture as a practical baseline for integrated cruise-to-park automated valet parking studies. Full article

Background/Objectives: Radial artery spasm (RAS) is an important complication during transradial coronary angiography that may negatively affect procedural success and reduce patient comfort. The aim of this study was to comparatively evaluate the effects of intravenous (IV) and intra-arterial (IA) heparin administration on the development of RAS. Methods: This prospective, observational parallel-group cohort study included a total of 223 patients undergoing transradial coronary angiography. Patients were divided into two groups, receiving either IV heparin (n = 77) or IA heparin (n = 146). All patients received a standard dose of unfractionated heparin (5000 IU) and an IA spasmolytic cocktail consisting of 2.5 mg verapamil and 100 mcg nitroglycerin. RAS was defined as pain during the procedure, resistance during catheter manipulation, or the need for crossover. Logistic regression analysis and receiver operating characteristic (ROC) curve analyses were performed. Results: RAS developed in 40 of 223 patients (17.9%). The incidence of RAS was significantly higher in the IA heparin group than in the IV heparin group (23.3% [34/146] vs. 7.8% [6/77]; p = 0.004). Crossover to femoral access due to severe spasm was observed only in the IA group (6.2% [9/146] vs. 0% [0/77]; p = 0.026). Patients who developed RAS were younger, required a greater number of catheters, had longer angiography duration, and were exposed to a higher total radiation dose (p < 0.05 for all). In ROC analysis, the number of catheters used and angiography duration showed comparable performance in predicting RAS. In multivariable logistic regression analysis, IA heparin administration and the number of catheters used were identified as independent predictors of RAS. Conclusions: During transradial coronary angiography, intravenous heparin administration is associated with a significantly lower frequency of RAS and a reduced need for femoral crossover compared with intra-arterial administration. IV heparin may represent an easily applicable strategy for RAS prevention, although causality cannot be established from this observational study. Full article

To address the problems of strong vertical heterogeneity in thin interbedded reservoirs of the N block in Daqing Oilfield, the complex coupling between porosity and permeability, and the difficulty of conventional single-parameter inversion methods in balancing local details with global geological background, a joint inversion method for porosity and permeability based on GeoFE-PPNet and logging imaging tensors is proposed. Using conventional logging curves, including GR, RT, RHOB, NPHI, DT, and PE, the method constructs a logging imaging tensor by integrating multi-channel responses with shale constraints and extracts intra-layer textural features through local encoding. Meanwhile, sequence decomposition and frequency enhancement are introduced to capture vertical trend variations and high-frequency non-stationary responses of the reservoir. On this basis, geological constraint fusion and dual-task collaborative prediction are employed to achieve joint inversion of porosity and permeability. Experimental results show that the proposed method achieves favorable inversion accuracy and cross-well generalization under complex reservoir conditions, with a porosity R² of 0.931, a permeability R² of 0.887, and an overall accuracy of 90.74%. Ablation and noise robustness experiments further demonstrate the effectiveness of the logging imaging tensor, frequency enhancement, geological constraints, and dual-task collaboration in improving model performance. The study indicates that the proposed method can more accurately characterize the vertical variation in reservoir physical properties and provides a new technical approach for fine reservoir evaluation and intelligent log interpretation. Full article