Search Results (2,722)

Background: Acromegaly is a rare, chronic, systemic, and progressive disease characterized by an excess secretion of growth hormone (GH) and increased circulating insulin-like growth factor 1 (IGF-1) concentrations, typically due to a macroadenoma in the pituitary gland. Both GH and IGF-1 are implicated in cancer promotion based on experimental and epidemiological data, but research findings remain conflicting and population-based data are scarce. Although there is a high mortality rate among acromegalic patients due to cardiovascular diseases, cancer is the third leading cause of death. Aim: The aim of the present study was to assess the risk of different types of cancer in acromegaly and the impact of changes in disease control and patient outcomes over time. Methods: Patients diagnosed with acromegaly at the Ankara University Ibn-i Sina Hospital Endocrinology and Metabolic Diseases Department between 2015 and 2019 were included in this study. Data including demographic data, history of cancer, size of adenoma (micro or macro), serum IGF-1 and GH levels at the time of diagnosis, serum prostate-specific antigen (PSA), thyroid ultrasonography, and, if needed, thyroid fine needle aspiration cytology (TFAC), colonoscopy, and mammography results were collected from patient records retrospectively. Results: We screened 83 patients, and 78 patients with the compensatory data (female/male: 39/39, 50%/50%) were included. The mean age of patients was 49.4 ± 11.9 years and 41.7 ± 12.1 years at the time of diagnosis. The median duration of follow-up was 72 (12–420) months. Periodic thyroid ultrasonography was performed in 65/78 (83.3%) of the patients, and a colonoscopy and mammography were also conducted in 27/78 (34.6%) and 32/39 (82%) of the patients at least once over the course of the disease, respectively. Cancer was detected in 17/78 (21.7%) of the patients; 11/78 (14.1%) of them had well-differentiated thyroid cancer and 2/39 (5.1%) had breast cancer. Prostate cancer, renal cell carcinoma, pancreatic cancer, malignant chordoma, schwannoma, and colon cancer were detected in one patient each. The increased cancer risk in acromegalic patients did not correlate with age, sex, age at diagnosis, time to diagnosing acromegaly, duration of acromegaly, GH and IGF-1 levels at diagnosis, pituitary adenoma size, or Ki-67 levels. Conclusions: Cancer was detected in 21.7% of the acromegaly patients, 14.1% of whom had well-differentiated thyroid cancer. In this study, we demonstrated that thyroid cancer is the most common malignancy in Turkish acromegalic patients, consistent with the results of previous studies. The increased cancer risk in acromegalic patients did not correlate with age, sex, age at diagnosis, time to diagnosing acromegaly, duration of acromegaly, or GH and IGF-1 levels at diagnosis. Full article

In recent years, cycling mobility in urban environments across Spain has grown significantly, driven by sustainability policies and behavioral shifts following the COVID-19 pandemic. However, this growth has been accompanied by an increase in accidents in urban areas, where more than 72.6% of cyclist accidents are concentrated, with large cities being the most affected. This study aims to explore and analyze the factors influencing cycling accidents in Spanish municipalities with populations exceeding 50,000, during the period of 2020–2023. A total of 24 variables were analyzed, encompassing not only innovative cyclist infrastructure network features (line connectivity), but also urban morphology and street infrastructure, weather conditions and mobility (all transportation modes). The methodological approach combines Principal Component Analysis (PCA) with two negative binomial regression models: one addressing all cycling accidents, and another focusing specifically on collisions between cyclists and motor vehicles. PCA shows the complex relations between urban features when comparing cyclist accidents among cities. The main results from the Negative Binomial analysis show that increased bicycle lane length significantly reduces cycling accident risk, while higher intersections with traffic signal density are associated with a greater likelihood of car–bicycle crashes. These findings emphasize the importance of cycling infrastructure provision and intersection design and regulation as key policy levers for improving urban cyclist safety. Future research should seek to corroborate these results through micro-spatial analyses and accident geolocation, assessing their severity and accounting for more detailed data on cycling infrastructure. Finally, the results’ discussion underscores the importance of implementing holistic urban mobility strategies that prioritize cyclist safety. Full article

The maturity of cannabis flowers at harvest critically influences cannabinoid yield and product quality. However, conventional assessment methods rely on subjective visual inspection of trichomes and stigmas, making them inherently inconsistent. This research presents an automated framework integrating computer vision and deep learning to objectively evaluate cannabis flower maturity. High-resolution macro images were acquired using low-cost smartphone-based systems under greenhouse and laboratory conditions. A two-stage pipeline was implemented: a fine-tuned Faster R-CNN model detected trichomes (Precision: 0.815; Recall: 0.802), while a YOLOv8 classifier categorized them into clear, milky, or amber classes (Accuracy: 98.6%). In parallel, a YOLOv8 segmentation model delineated stigmas (AP50: 52.2%) to compute color ratios as maturity indicators. Features were aggregated at the flower level and correlated with HPLC-measured cannabinoid concentrations. A dataset of over 14,000 images was collected across multiple imaging sessions to support training, evaluation, and correlation experiments. Results demonstrated that stigma coloration—detectable with low-end devices—provides a robust visual indicator of peak chemical maturity, with the green-to-orange transition aligning with maximum cannabinoid concentration. This work offers a scalable, cost-effective solution for real-time maturity assessment in cannabis cultivation, contributing to improved harvest timing and quality control. Full article

At least seven species of wild cats inhabit Colombia, and these species are also distributed throughout Mexico, Central America, and the rest of South America (jaguar, puma, jaguarundi, ocelot, margay, tigrina, and Pampas cat). A checklist and review of phylogeographic and population genetic studies on these seven wild cat species has been conducted here, as this information is vital for effective conservation programs. The jaguar is the feline species on which the most population genetic studies have been carried out in the Neotropics. In general, little genetic structure has been found at a macro-geographic scale. Genetic diversity is high in countries such as Colombia, Peru, and Bolivia, and generally throughout the Amazon basin. However, genetic diversity is more moderate or even significantly lower in Mexico and the Brazilian Atlantic Forest. Much of the genetic research on the jaguar has focused on Brazil, Mexico, and Belize, but Colombia is also very well represented in these studies. However, there is a complete or very pronounced lack of data in other areas such as Venezuela, the Guianas, some Central American countries, Paraguay, and northern Argentina. After the jaguar, the most studied feline in Neotropics from a population genetics perspective is the puma. In North America, this species has low genetic diversity, while the diversity in Central America is moderate, and South America is where genetic diversity is highest. The countries best represented in these studies are Brazil (southern of the country), Mexico, Belize, and Argentina. However, countries like Colombia, Ecuador, Peru, and Paraguay are very poorly represented in puma genetic studies. Very few genetic studies have been conducted on the jaguarundi, despite its vast geographic distribution. In northern Mexico, its genetic diversity is very low, but in countries like Colombia, Peru, and Bolivia, genetic diversity is very high. Colombia is probably the country where jaguarundis have been genetically studied most extensively. The third wild cat species with the most molecular studies in Neotropics is the ocelot, although it lags considerably behind jaguars and pumas. Its genetic diversity is low in Texas and northern Mexico, but very high, especially in countries surrounding the Amazon basin. A good number of macro-geographic studies have been conducted on the ocelot, and these studies are very representative of ocelots in countries such as Colombia (probably the best represented country), Ecuador, Peru, Bolivia, Panama, and Costa Rica. However, there are other countries where molecular studies of the ocelot have not been carried out, such as Paraguay and Argentina, with the lack of such studies in Brazil being particularly noteworthy. Very few molecular studies have been conducted on the margay. In general, its genetic diversity is very high in all the studies performed. Colombia, Peru and Bolivia are very well represented, but the lack of molecular studies in Mexico, much of Central America, and South American countries such as Brazil, Paraguay, and Argentina is striking. The tigrina is one of the Neotropical wild cat species that requires the most molecular studies to unravel its complex systematics. Only the southern Brazilian tigrina (Leopardus guttulus), which was elevated to a new species, has generated comprehensive molecular information. Molecular studies of the Andean tigrina have revealed a very complex picture that, at present, does not allow us to know exactly how many species or taxa inhabit that area of South America and, therefore, to develop a conservation program that adequately correlates with that number of taxa. Finally, in the case of the Pampas cat, molecular studies are well represented with specimens from Peru, Bolivia, Chile, Argentina, Brazil and Uruguay. Studies are needed in Paraguay, and especially in Ecuador and southern Colombia (assuming a stable population of Pampas cats exists in the latter country), where even at the molecular level, the specific taxon (one species or five species) present has not yet been determined. All this information is essential for developing effective regional and global conservation programs for these magnificent creatures. In Colombia, the development of molecular studies with the puma, the tigrina, and the Pampas cat is of special interest. Full article

The COVID-19 pandemic exposed significant vulnerabilities in public-sector administrative capacity, particularly in the implementation of crisis-related state aid schemes. Under conditions of extreme workload, time pressure, and legal constraints, administrative processes became critical determinants of policy effectiveness rather than routine implementation mechanisms. This study examines how such processes perform under crisis conditions and whether process modeling and simulation can identify efficiency gains without undermining procedural control. Using a case study of a COVID-19 state aid scheme administered by the Ministry of Transport of the Slovak Republic, the study combines Business Process Model and Notation (BPMN)-based process modeling, discrete-event simulation, and Monte Carlo analysis, and can identify efficiency gains in crisis-related state aid administration. The methodological approach integrates BPMN-based process modeling, discrete-event simulation, and scenario-based (“what-if”) sensitivity analysis to evaluate process performance under crisis-induced demand surges. Key performance indicators, including processing time, labor costs, and resource utilization, are analyzed using simulation outputs and dashboard-based visualization. Data analysis is conducted through simulation-based evaluation of key performance indicators, including processing time, labor costs, queue length, and resource utilization, under both baseline (AS-IS) and redesigned (TO-BE) process configurations. Scenario-based (“what-if”) and sensitivity analyses are applied to assess the effects of crisis-induced demand surges and capacity constraints on administrative performance. The results show that increased application volume during the crisis led to disproportionate growth in processing times due to queue accumulation and resource contention. Simulation-based process redesign reduced the average process cycle time by up to 12.8% and labor costs per application by up to 8.4% compared to the AS-IS configuration. However, efficiency gains diminished as resource utilization approached capacity limits, indicating structural constraints inherent to public administration. These findings demonstrate that process-oriented simulation provides a robust analytical tool for understanding administrative behavior under crisis conditions and for designing more efficient and resilient state aid mechanisms. The study contributes to public administration research by offering a micro-level, process-based perspective on crisis governance that complements the existing macro-level policy evaluations. Full article

Histopathological image analysis remains the cornerstone of cancer diagnosis; however, manual assessment is challenged by stain variability, differences in imaging magnification, and complex morphological patterns. The proposed multi-pretrained deep learning fusion (MPDLF) approach combines two widely used CNN architectures: ResNet50, which captures deeper semantic representations, and VGG16, which extracts fine-grained details. This work differs from previous fusion studies by providing a controlled evaluation of early, intermediate, and late fusion for integrating two pretrained CNN backbones (ResNet50 and VGG16) under single-modality histopathology constraints. To isolate the fusion effect, identical training settings are used across three public H&E datasets. Early fusion achieved the best test performance for the two primary tasks reported here: breast cancer binary classification (accuracy = 0.9070, 95% CI: 0.8742–0.9404; AUC = 0.9707, 95% CI: 0.9541–0.9844) and renal clear cell carcinoma (RCCC) five-class grading (accuracy = 0.8792, 95% CI: 0.8529–0.9041; AUC (OvR, macro) = 0.9895, 95% CI: 0.9859–0.9927). Future work will extend these experiments to additional magnification levels (100×, 200×, and 400×) for breast cancer histopathology images and explore advanced hybrid fusion strategies across different histopathology datasets. Full article

Seed priming studies commonly emphasize growth and physiological responses, yet ionomic regulation and tissue-specific nutrient allocation under salinity stress remain poorly explored, particularly in underutilized crops such as Bambara groundnut (Vigna subterranea L.). This study investigated whether Mg(NO₃)₂ seed priming, previously shown to enhance salt tolerance, is associated with consistent ionomic patterns in contrasting Bambara groundnut genotypes (BGN-14 and BGN-25). Seeds were primed with 0.03% Mg(NO₃)₂ and grown under control or saline conditions (200 mM NaCl) for five weeks. Shoot and root tissues were analyzed for macro- and micronutrient composition using ICP-OES. In BGN-14, salinity caused a marked reduction in shoot fresh weight (−49.5%, p < 0.05), whereas Mg(NO₃)₂ priming largely mitigated this effect under salinity (−0.4%, p > 0.05). Root fresh weight declined numerically under salt stress (−70.1%) and primed + salt conditions (−45.5%), but these changes were not statistically significant. Shoot dry weight increased significantly in primed plants (+83.5%, p < 0.05), while salinity reduced SDW (−58.4%); primed + salt plants maintained SDW near control levels (+2.6%). In BGN-25, root biomass was unaffected by treatments, whereas salinity significantly reduced shoot biomass relative to primed plants, with a consistent trend of primed > control > primed + salt > salt. Salinity increased the Na⁺/K⁺ ratio, particularly in roots. In BGN-14, the root Na⁺/K⁺ ratio increased significantly from 1.07 to 4.49 (p < 0.05), indicating enhanced Na⁺ accumulation, while shoot ratios increased non-significantly. BGN-25 showed a more moderate increase in shoot ratios and a pronounced rise in root ratios. Principal component analysis revealed distinct nutrient clustering, with Na, Fe, and Al loading strongly under salinity, while Ca, K, Mg, and Cu aligned with improved physiological performance. Although differences between salt and primed + salt treatments were often not statistically significant, several ion ratios and nutrient relationships were numerically enhanced under Mg(NO₃)₂ priming. This study builds upon earlier physiological findings (where BGN-14 consistently exhibited a stronger positive response to Mg(NO₃)₂ priming, outperforming BGN-25 under salt stress) and provides exploratory, hypothesis-generating evidence that Mg(NO₃)₂ priming may contribute to salinity tolerance through coordinated ionomic adjustments, including altered Na⁺ allocation and improved nutrient balance, rather than complete Na⁺ exclusion. These findings highlight the relevance of ionomic responses in understanding stress adaptation in underutilized legume crops. Full article

Background: Boundary-spanning bank employees experience continuous customer interactions that can generate negative emotions and influence how they perceive interpersonal treatment at work. Although emotions are known to shape fairness judgments, little is understood about how negative emotions translate into lower interactional justice through self-related psychological mechanisms. This study examines whether organization-based self-esteem (OBSE), self-efficacy, and external workplace locus of control (EWLOC) jointly condition the association between negative emotions and interactional justice. Method: Survey data were obtained from 338 boundary-spanning bank workers in Istanbul, Türkiye. The proposed moderated mediation framework was examined using PROCESS Macro Model 21, testing OBSE as a mediator between negative emotions and interactional justice. Self-efficacy and EWLOC were included as moderators on the first- and second-stage paths. Bootstrapping with 5000 resamples provided confidence intervals for the conditional indirect effects. Results: Stronger negative emotions were associated with lower OBSE and lower interactional justice perceptions, and the pattern of results was consistent with an indirect association via OBSE. The negative emotion–OBSE association varied by self-efficacy, with the association being stronger among employees reporting lower self-efficacy. EWLOC conditioned the OBSE–justice association, such that this association was weaker at higher levels of external orientation. The conditional indirect association was significant for employees low in self-efficacy and low-to-moderate in EWLOC. Conclusions: The findings are consistent with a moderated mediation model in which negative emotions are indirectly linked to interactional justice through OBSE, with the magnitude of these associations depending on employees’ personal resources and attributional tendencies. The results suggest that strengthening OBSE and related psychological resources may be useful for supporting justice perceptions in emotionally demanding service roles. Full article

Visibility is a critical meteorological factor for ensuring the safety of maritime and bridge transportation, and accurate identification of low-visibility levels is essential for early warning and operational scheduling. Traditional methods such as Random Forest often exhibit insufficient feature-modeling capability when dealing with high-dimensional, multi-source remote sensing data. Meanwhile, satellite observations used for visibility recognition are characterized by strong inter-channel correlations, complex nonlinear interactions, significant observational noise and outliers, and the scarcity of low-visibility samples that are easily confused with low clouds and haze. As a result, existing general deep learning methods (e.g., the Saint model) may still exhibit unstable attention weights and limited generalization under complex meteorological conditions. To address these limitations, this study constructs a visibility classification task for the Jiaxing–Shaoxing Cross-Sea Bridge region in China based on multi-channel visible and infrared spectral observations from the Fengyun-4A (FY-4A) and Fengyun-4B (FY-4B) satellites. We propose a visibility classification method using the LF-Transformer for the Jiaxing–Shaoxing Cross-Sea Bridge region in China, and systematically compare it with the Random Forest and Saint models. Experimental results show that the Precision of the LF-Transformer increases significantly from 0.47 (Random Forest) to 0.59, achieving a 13% improvement and demonstrating stronger discriminative ability and stability under complex meteorological conditions. Furthermore, a combination input of FY4A+FY4B outperform the single FY4A, with a 25.5% increased Macro F1-score. With an additional ensemble strategy, the LF-Transformer further improves its precision on the FY4A+FY4B fused dataset to 0.61, a 3% compared to the original LF-Transformer, indicating enhanced prediction stability. Overall, the proposed method substantially strengthens visibility classification performance and highlights the strong application potential of the LF-Transformer in remote-sensing-based meteorological tasks, particularly for low-visibility monitoring, early warning, and transportation safety assurance. Full article

Background: Occupational gender segregation is a key driver of labor market inequality and is prominent across occupations, within occupations, and within workplace task groups. This paper explores how structural arrangements and cultural messages shape women’s preferences for working with men vs. women. With respect to structural arrangements, we analyze how women’s relative power on a team influences their partner preference. With respect to cultural messages, we examine how one common source of information that has the potential to either challenge or reify notions of gender difference—sexual harassment policy training—affects partner preference. Methods: We conducted a laboratory experiment in which we placed 100 college-aged women in positions they may commonly find themselves in at the start of a new job—identifying coworkers to partner with on group tasks—and varied (1) their relative power on the team (leader or helper) and (2) exposure to workplace training (sexual harassment or ergonomic computer setup). We then assessed their attitudinal and behavioral preference for working with a female vs. a male partner on a decision-making task. Results: Women, particularly women assigned to a leadership position, more often chose to work with a female partner. Sexual harassment training did not affect women leaders’ attitudes about working with a male partner but those in a helper role expressed more positive attitudes about working with a man after sexual harassment training. These findings document how macro-level processes can shape workplace gender segregation, thus identifying mechanisms underlying the reproduction of gender inequality. Full article

Cross-domain sentiment analysis remains challenging due to distributional shifts and heterogeneous sentiment expressions across platforms. Existing domain adaptation approaches primarily focus on enforcing domain-invariant representations. However, such symmetry-preserving strategies often overlook directional and expression-level asymmetries. These asymmetries naturally arise in real-world sentiment data, particularly for context-inferred sentiment expressions. In this work, we propose a novel symmetry- and asymmetry-aware domain adaptation framework for cross-domain sentiment classification. The framework models symmetry through explicit multi-source distribution alignment, which captures transferable sentiment knowledge across domains. Additionally, aspect-level structural supervision organizes representations according to shared linguistic aspects. To address asymmetry, a directional divergence regularization is introduced. This component models expression-level and directional discrepancies between source and target domains. Importantly, the framework operates without requiring target-domain annotations. Experiments are conducted under a multi-source unsupervised domain adaptation setting using sentence-level hotel review datasets collected from multiple online platforms. Empirical results demonstrate strong performance for the proposed framework. It achieves an average Accuracy of 82.0% and Macro-F1 of 80.6%. The framework consistently and statistically significantly outperforms source-only, multi-source, and transformer-based adversarial adaptation baselines across all evaluated target domains (p < 0.05). Additional analyses confirm improved robustness to implicit sentiment expressions and platform-induced asymmetries. These findings highlight the importance of jointly modeling symmetry and asymmetry for robust cross-domain sentiment adaptation and provide a unified and deployable solution for sentiment analysis under realistic platform shifts. Full article

Background: Obstructive Sleep Apnea (OSA) is highly prevalent yet underdiagnosed due to the scarcity of Polysomnography (PSG) resources. Audio-based screening offers a scalable solution, but often lacks the granularity to precisely localize respiratory events or accurately estimate the Apnea-Hypopnea Index (AHI). This study aims to develop a fine-grained and lightweight detection framework for OSA screening, enabling precise respiratory event localization and AHI estimation using non-contact audio signals. Methods: A Dual-Stream Convolutional Recurrent Neural Network (CRNN), integrating Log Mel-spectrograms and energy profiles with BiLSTM, was proposed. The model was trained on the PSG-Audio dataset (Sismanoglio Hospital cohort, 286 subjects) and subjected to a comprehensive three-level evaluation: (1) frame-level classification performance; (2) event-level temporal localization precision, quantified by Intersection over Union (IoU) and onset/offset boundary errors; and (3) patient-level clinical utility, assessing AHI correlation, error margins, and screening performance across different severity thresholds. Generalization was rigorously validated on an independent external cohort from Beijing Tongren Hospital (60 subjects), which was specifically curated to ensure a relatively balanced distribution of disease severity. Results: On the internal test set, the model achieved a frame level macro F1 score of 0.64 and demonstrated accurate event localization, with an IoU of 0.82. In the external validation, the audio derived AHI showed a strong correlation with PSG-AHI (r = 0.96, MAE = 6.03 events/h). For screening, the model achieved sensitivities of 98.0%, 89.5%, and 89.3%, and specificities of 88.9%, 90.9%, and 100.0% at AHI thresholds of 5, 15, and 30 events per hour, respectively. Conclusions: The Fine-Grained and Lightweight Dual-Stream CRNN provides a robust, clinically interpretable solution for non-contact OSA screening. The favorable screening performance observed in the external cohort, characterized by high sensitivity for mild cases and high specificity for severe disease, highlights its potential as a reliable tool for accessible home-based screening. Full article

Amid growing international pressure for carbon neutrality, the maritime industry is facing mounting institutional demands for environmental innovation. Drawing on protection motivation theory, this study surveyed 499 employees from 1519 shipping service firms to examine how coercive, mimetic, and normative pressures shape green innovative work behavior. By extending protection motivation theory to a highly regulated maritime context, this study demonstrates that institutional pressures do not directly drive green innovation; instead, they enhance employees’ self-protective motivation, which subsequently fosters eco-innovation. Moreover, these relationships are stronger when firms perceive greater climate risks or receive government subsidies, indicating that contextual conditions amplify the translation of motivation into behavior. Overall, the findings reveal how macro-level institutional forces shape the sustainable transition of maritime services through employee psychology, offering governance-relevant insights for policymakers and firms seeking to promote green innovation. Full article

To address the persistent challenges of the “disconnect between macro-level spatial zoning and micro-level land allocation” and the paradox of “localized intensification accompanied by overall inefficiency” in territorial spatial governance, this study adopts a multi-scale coupling perspective to explore an optimized spatial pattern that promotes the coordinated development of production, living, and ecological functions (PLEFs), thereby enhancing the systematic and scientific basis of spatial governance. Taking the Central Yunnan Urban Agglomeration (CYUA) as a case study, a coupled optimization framework integrating macro-scale spatial zoning and micro-scale land allocation was established. First, using multi-period land use data (2000–2020) in conjunction with multi-source geographic and socio-economic datasets, the correspondence between land use types and PLEFs was constructed, and the spatiotemporal evolution characteristics of these functions were systematically analyzed. Second, the GMOP-PLUS model was employed to simulate and optimize land use patterns for 2035 under multiple development scenarios, and dominant spatial types were further refined based on grid-scale spatial suitability evaluation results. Third, the NRCA model was applied to identify comparative functional advantages at the county level. These advantages were comprehensively integrated with the revised dominant spatial types following the principle of “seeking common ground while preserving differences,” ultimately enabling the delineation of optimized territorial spatial development zones. The results indicate the following: (1) Throughout the study period, ecological space remained the dominant land use type (exceeding 75%), although its proportion gradually declined. Living space expanded markedly, while the internal structure of production space shifted toward industrial and mining land uses. (2) The planning control scenario was identified as the most suitable development pathway, exhibiting a spatial configuration characterized by a “central core with stronger development in the southeast than in the northwest.” Under this scenario, production and living spaces continued to expand, whereas ecological space maintained its dominant status. (3) Spatial suitability evaluation revealed a high degree of functional compatibility across the study area (79.01%), with ecological suitability prevailing. The revised dominant spatial types were predominantly ecological (78.94%), forming a hierarchical structure described as a “living core–production composite ring–ecological periphery.” (4) By integrating dominant spatial types with comparative functional advantages, the study area was classified into five major categories of territorial spatial development zones, for which differentiated governance strategies were proposed. This study provides methodological insights and practical guidance for improving refined territorial spatial management and advancing regional sustainable development. Full article

Humanitarian crises such as natural disasters and armed conflicts are increasing in frequency and intensity, posing major challenges to the sustainable protection of vulnerable populations. Rapid and equitable identification of individuals at highest risk is essential for efficient allocation of limited emergency resources and for strengthening community resilience. This study proposes an intelligent, privacy-aware decision-support platform for citizen-level vulnerability assessment that supports social sustainability and resilient crisis management. The platform integrates heterogeneous data from healthcare institutions, municipal civil records, and emergency rescue services to construct multidimensional vulnerability profiles based on social conditions, medical status, and geographical accessibility. The dataset was collected in Algeria in collaboration with the Algerian Civil Protection and consolidated into a fully anonymised dataset of approximately 5000 individual records reflecting realistic crisis scenarios. Five supervised machine learning models (Decision Tree, Random Forest, Support Vector Machine (RBF), XGBoost, and Logistic Regression) were evaluated under class-imbalance conditions using SMOTE and class weighting. The Random Forest model achieved the best performance, with an F1-Macro score of 0.710 and a recall of 0.569 for the high-risk class (95% confidence interval: [0.431, 0.706]). These results demonstrate that the proposed platform enables transparent, data-driven prioritisation of emergency interventions, contributing to sustainable humanitarian response, improved public resource allocation, and enhanced resilience of vulnerable communities. Full article