Search Results (2,070)

This study examines the spatiotemporal evolution and driving mechanisms of urban expansion in the Guangxi Zhuang Autonomous Region, China, from 2013 to 2023. Using Suomi-NPP VIIRS nighttime light (NTL) data, we combine Standard Deviational Ellipse (SDE) analysis, centroid migration, kernel density estimation (KDE), landscape metrics, Local Moran’s I (LISA), and system Generalised Method of Moments (system-GMM) estimation. The results show that the centroid of urban development remained within Binyang County while moving overall toward the southeast with recurrent north–south oscillations. The SDE results indicate a stable northeast–southwest orientation, with secondary expansion in other directions. The urban structure is dominated by a strong Nanning core, accompanied by secondary clusters in Liuzhou, Guilin, and other prefecture-level cities. Nanning recorded the largest absolute expansion, followed by secondary centres, including Liuzhou, Guilin, Yulin, Wuzhou, Fangchenggang, Qinzhou, and Beihai, whereas western and northern Guangxi expanded more slowly. The system-GMM results indicate that financial deepening has a marginally significant positive effect on built-up area expansion and fiscal pressure has a marginally significant constraining effect, both at the 10% level; land finance dependency does not emerge as an independent driver in this small panel. We interpret these findings through a Source–Channel–Valve framework, in which financial deepening provides the capital source, land finance represents a hypothesised institutional channel, and fiscal pressure acts as a regulatory constraint. The study provides empirical evidence for sustainable and regionally coordinated urban development in Guangxi and comparable geographically constrained regions. Full article

This study examines the association between foreign exchange (FX) governance and financial stability by analysing empirical evidence from multinational entities. We analyse a 16-year panel (2009–2024) comprising 6613 firm-year observations using OLS regression with industry and year fixed effects. Firm-level data on financial sustainability, FX governance, board attributes, and controls are drawn from the London Stock Exchange Group (formerly Refinitiv), while country-level institutional and economic indicators are obtained from the World Bank. The result suggests that FX governance is negatively associated with earnings volatility, implying that FX governance enhances the financial stability of organisations. The baseline result is robustness to endogeneity and selection bias. However, our subsample analysis reveals that the impact of FX governance on financial stability varies based on institutional quality and industry. Whereas FX governance is negatively associated with earnings volatility thus enhancing financial stability in high-institutional-quality settings, the impact is not significant in low-institutional-quality environments. This study contributes to knowledge by empirically validating the relevance of FX governance to financial stability. Our study also contributes to the limited studies on the role of FX governance in diminishing earnings volatility, thus exposing FX management as a strategy for achieving financial sustainability. The international sample analysed in the study contributes to the generalisability of results. Full article

Rift Valley fever (RVF) is a viral zoonosis endemic in Madagascar, threatening human and animal health as well as the economy. Trade-related livestock movements are a major factor in the spread of RVF virus. While previous RVF research in Madagascar has focused on farmers or general ecology, this study is the first to specifically target livestock traders, the primary drivers for long-distance viral spread, in the Alaotra Mangoro endemic hotspot. This study aimed to assess the level of knowledge, prevailing attitudes and current practices regarding RVF among people engaged in livestock trade in the Alaotra Mangoro region, as well as the factors associated with these KAPs. A descriptive and analytical cross-sectional survey was conducted among 406 livestock traders in five districts of the Alaotra Mangoro region, using a structured questionnaire. A multi-stage sampling approach was employed, utilising purposive selection of markets followed by snowball sampling to reach informal traders often missed by traditional surveys. Generalised linear mixed models were used to analyse factors associated with KAPs regarding RVF. Awareness of RVF was very low (only 18.5% respondents had heard of it), with significant regional disparities (0% in Anosibe An’Ala versus 51.6% in Moramanga). Veterinarians (15.5%), family (12.8%), radio (9.6%) and neighbours (9.6%) were the main sources of information. Understanding of symptoms and modes of transmission (particularly mosquito bites) was limited. Higher levels of education (OR = 181.6; 95% CI: 29.9–1123.7; p < 0.001) and older age (50–60 years) were associated with better knowledge. Proactive attitudes were scarce (21.4%), although more than half (53.4%) believed that RVF is a real disease. Perception of personal risk and the contribution of livestock trade to the spread of the disease was low. However, confidence in animal vaccination was relatively high (60.3%). Preventive practices were highly inadequate. The majority did not wear protective equipment when handling sick animals (94.6%) and rarely avoided touching aborted foetuses (12.6%). Less than half (48.3%) expressed a willingness to report sick or dead animals, and nearly half admitted to having sold or purchased sick livestock (49.5%). Cooking meat (95.1%) and using mosquito nets (74.1%) were the only well-established practices. More than half of respondents (57.9%) lived more than 5 km from veterinary services, and cost was the most frequently cited barrier to consultation. Participation in awareness campaigns was virtually non-existent (5.4%). Results revealed critical gaps in KAP that may contribute to the persistence of RVF. A “One Health” approach is imperative, integrating human, animal and environmental health. Full article

The growing integration of rehabilitation robotics and artificial intelligence has created new opportunities for developing control strategies that better support clinicians during patient therapy. This study investigates machine learning and deep learning approaches for classifying upper limb motion using encoder-based biomechanical data, with the goal of identifying a model suitable for implementation in a rehabilitation exoskeleton. Several classical algorithms such as k-Nearest Neighbors, Random Forest, multiclass logistic regression, XGBoost, and an SVM classifier were evaluated alongside three deep learning architectures: convolutional layers, GRU and LSTM units. Models were trained and tested on two types of datasets using both standard cross-validation and leave-one-subject-out validation. The analysis included assessments of class separability, signal features’ importance, and comparative performance based on F1-score, accuracy, and confusion matrices. Results showed notable differences between validation strategies, with LOSO evaluation revealing limitations of the available dataset and emphasising the need for broader data collection. Overall, the findings indicate that, in the LOSO evaluation of the five-class multi-subject dataset—the most clinically realistic validation scenario—the LSTM-based model achieved the highest generalisation performance (accuracy 92.8%, macro-F1 0.927), supporting its suitability for integration into exoskeleton control systems aimed at detecting and mitigating compensatory movements. Full article

This paper investigates the relationship between spatial openness and cognitive engagement, integrating geometric and neurophysiological indicators to address the lack of frameworks directly coupling spatial structure with neural responses. Spatial openness is quantified using three-dimensional isovist volume. Engagement is measured via an EEG-based index (β/(θ + α)). Twenty-six participants completed an experiment in a virtual reality environment in which 16 spatial conditions of varying openness were presented. A node-based framework couples spatial metrics with EEG responses at the level of individual observation points and temporal segments. Linear and quadratic mixed-effects models reveal a small but statistically detectable inverted-U relationship between openness and engagement (marginal R² = 0.020) that persists after correction for spatial–temporal autocorrelation, with the pattern replicated in 18 of 26 participants. We interpret these findings as preliminary neurophysiological evidence that spatial openness modulates engagement through an optimal range of stimulation, supporting designs that balance visual exposure against spatial boundaries. Generalisation is constrained by the VR-based setting, the limited sample size, and the small absolute effect. Full article

Sepsis is a prominent cause of death in intensive care units, and delayed diagnosis greatly worsens fatal outcomes due to the complex, irregular, and uneven character of clinical time-series data. Hence we proposed an interpretable and uncertainty-aware deep learning architecture that solves data quality, temporal irregularity, and clinical explainability restrictions, which are frequently addressed separately by existing models. The suggested method combines Bidirectional Recurrent Imputation for Time Series (BRITS)-based imputation, hybrid Conditional Tabular Generative Adversarial Network-Synthetic Minority Over-sampling Technique (CTGAN-SMOTE) data augmentation, a Temporal Convolutional Networks (TCN)-Attention architecture, and continuous-time neural Ordinary Differential Equations (ODEs), along with SHapley Additive exPlanations (SHAP)-based feature attribution and uncertainty quantification. The experimental evaluation on a large ICU dataset reveals greater predictive accuracy, with an AUROC of 0.926 and accurate early warnings up to six hours before clinical onset, all while maintaining strong interpretability and calibration. The proposed framework demonstrates strong predictive performance and provides early warnings up to six hours before clinical onset, while maintaining interpretability and calibration. While the results are promising, further validation across multiple clinical settings is required to confirm its generalisability and real-world applicability. Full article

Reliable tracking of multi-pollutant air-quality progress is essential for assessing policy effectiveness and health risks, yet most assessments still focus on single pollutants. We analysed population-weighted exposures to fine particulate matter (PM₂.₅), nitrogen dioxide (NO₂) and household air pollution (HAP) for Australia and the global average over 1990–2019, using harmonised estimates from a Global Burden of Disease–type framework. Non-parametric LOESS and semi-parametric generalised additive models were applied to characterise long-term trends, and a composite clean-air progress index (CAPI; 1990 = 1) was constructed to summarise joint changes in the three pollutants. Statistical and Monte Carlo methods were used to propagate reported exposure uncertainty into both pollutant-specific trends and the composite index. Globally, exposures to PM₂.₅, NO₂ and HAP all declined, and the CAPI fell to around 0.7 by 2019, indicating substantial multi-pollutant improvement relative to 1990. In Australia, NO₂ decreased more rapidly than the global mean, but PM₂.₅ showed little long-term decline and the HAP-related metric increased more than three-fold. As a result, Australia’s CAPI rose to approximately 1.6–1.7, with Monte Carlo uncertainty envelopes remaining well above 1 from the early 2000s onwards. Correlation analyses revealed that pollutants improved together at the global scale, but were partially decoupled in Australia, implying that source-specific gains have not translated into aggregate clean-air progress. These findings demonstrate that single-pollutant assessments can obscure important trade-offs and that multi-pollutant, uncertainty-aware indices such as CAPI provide a more informative basis for benchmarking national trajectories against global experience and for guiding integrated clean-air policy. Full article

Background/Objectives: During the first COVID-19 lockdown, the sudden disruption of formal care services substantially increased reliance on informal caregiving. Emerging evidence suggests that increased caregiving demands may have contributed to a higher burden of mental-health-related temporary work disability; however, population-based data from occupational health systems remain limited. This study aimed to quantify and characterise, descriptively, the sociodemographic, clinical, and territorial characteristics of mental-health-related temporary work disability among workers with informal caregiving responsibilities in Spain during the first COVID-19 lockdown, and to descriptively examine differences between episodes occurring among workers with and without caregiving responsibilities. Methods: A retrospective descriptive study was conducted using anonymised nationwide occupational health records from Mutua Asepeyo. All episodes of temporary work disability certified for mental and behavioural disorders (ICD-10 F00–F99) between 14 March and 21 June 2020 were analysed. Caregiver status was determined based on documented informal caregiving responsibilities recorded within the occupational disability records. Sociodemographic, occupational, clinical, and territorial variables were examined using descriptive statistics and non-parametric tests. Results: A total of 2857 caregiver-associated episodes were identified, representing 55.6% (95% CI: 54.2–57.0) of all mental-health-related temporary work disability episodes during the study period. The majority involved women (68.1%) and caregivers of older dependent adults (59.3%). Generalised anxiety disorder was the most frequent diagnosis, followed by adjustment disorders and acute stress reactions, with significant differences by sex and employment regime. Marked territorial variability was observed, as well as longer durations of temporary work disability in specific regions and among women. Conclusions: A substantial proportion of mental-health-related temporary work disability episodes during the lockdown occurred among workers with informal caregiving responsibilities, particularly women and those caring for older dependents. These findings suggest that informal caregiving may be a determinant of occupational mental health during crises. However, given the descriptive and unadjusted nature of the study, no causal inferences can be drawn. Further research is needed to understand these associations better and inform future occupational health strategies. Full article

Background/Objectives: Failure to rescue deteriorating ward patients before irreversible organ injury remains a leading cause of preventable in-hospital mortality, yet current rapid response team (RRT) research relies predominantly on cross-sectional comparisons at the moment of activation, overlooking the short-horizon physiological changes that precede it. Methods: This retrospective cohort study at a tertiary academic hospital in South Korea included 549 adults (191 ICU-transferred, 358 ward-remaining) with a first RRT activation between September 2023 and August 2025. Generalised estimating equations (GEE) with a time × group interaction modelled differential changes in 12 laboratory variables and the DeepCARS AI-derived risk score between 24 h before activation (T−24 h) and the moment of activation (T0). At T−24 h, physiological profiles were largely similar between groups, indicating that conventional static assessment failed to identify patients destined for ICU transfer. Results: Over the ensuing 24 h, patients subsequently transferred to the ICU showed a steeper decline in SpO₂/FiO₂ (S/F) ratio (383.4 → 167.1 vs. 369.1 → 260.3; B = −0.547, p < 0.001) and steeper increases in lactate (2.91 → 4.02 vs. 2.05 → 2.98 mmol/L; B = 0.154, p = 0.045), creatinine (B = 0.076, p = 0.038), potassium (B = 0.019, p = 0.001), and DeepCARS score (B = 0.073, p = 0.028) compared with patients remaining on the ward. All five variables retained significance under Benjamini–Hochberg false discovery rate correction (q < 0.10). Seven inflammatory and haematological markers showed no differential change. Procalcitonin was excluded from the primary analysis because of very high missingness at the pre-activation time point (approximately 75%). Conclusions: These findings demonstrate that short-horizon deterioration in oxygenation, perfusion, and renal function—rather than any single earlier measurement—distinguishes patients requiring ICU transfer, supporting the development of change-based early warning criteria to enable earlier clinical escalation. Full article

The call for a dialogue between the disciplines of anthropology and theology was initiated by anthropologist Joel Robbins in 2006. Within theology it was elaborated in 2014 by Michael Banner. This article compares both authors in order to understand what the highly generalising formulation of a dialogue between disciplines is about. They will turn out to aim at a conversation about what the disciplines are ultimately concerned with, formulated as otherness and everydayness, respectively. However, Robbins and Banner do not elaborate on their grand claims in a systematic and detailed way. This article offers a more systematic elaboration and aims to evaluate their views as a contribution to what is at stake in their calls for dialogue. For this purpose, it is necessary to better account for the complexity of the specific character of studying morality. Full article

In many African nations, including Nigeria, groundwater remains the most readily available source of clean water. However, finding and developing these resources in heterogeneous terrain, such as the Federal Capital Territory (FCT), Abuja, is challenging due to the uneven distribution of the aquifers and complex geological settings. Using a GIS-based machine learning approach that incorporates surface and subsurface hydrogeological parameters, this study defines groundwater potential zones (GWPZ). Nine conditioning factors were derived from remote sensing, geophysical and climatic datasets. Aquifer thickness, depth to bedrock, geology, rainfall, slope, LULC, lineament density, drainage density and distance from river were among these variables. Three machine learning models: Extreme Gradient Boosting (XGBoost), Support Vector Machine (SVM) and Random Forest (RF) were trained and validated using 2410 borehole records (productive and abortive). Hold-out validation (80:20), 10-fold cross-validation, ROC-AUC, and confusion matrix were used to assess each model’s performance. The ensemble models outperformed the SVM, achieving higher predictive accuracy and better generalisation (XGBoost: 0.89, RF: 0.88 and SVM: 0.87). The generated maps categorised the study area into five GWPZs: very high, high, moderate, low and very low. These findings provide a scientific foundation for groundwater exploration and sustainable water resource management in the study area. Full article

Background: Unscheduled bleeding in postmenopausal women on hormone replacement therapy (HRT) represents a common but poorly characterised clinical challenge. Whilst endometrial cancer affects approximately 9% of women with unexplained postmenopausal bleeding, existing evidence in HRT users is largely restricted to women under 60 years and short-duration regimens, leaving a critical evidence gap in contemporary all-age clinical practice. Whether the same investigative urgency is warranted for HRT users experiencing unscheduled bleeding as for women with unexplained postmenopausal haemorrhage remains unresolved. Objectives: To determine the diagnostic yield of endometrial cancer and hyperplasia amongst postmenopausal women presenting with unscheduled bleeding whilst on HRT, and to explore potential associations with HRT regimens and clinical risk factors. Methods: This retrospective observational study analysed 343 postmenopausal women presenting with unscheduled bleeding whilst on HRT at a single tertiary centre between September 2023 and February 2024. All patients underwent transvaginal ultrasound and endometrial sampling according to institutional protocol. Descriptive statistics were used to characterise outcomes, with exploratory analyses of potential risk factors. Given the symptomatic and selected nature of this cohort, all proportions represent the diagnostic yield within an investigated population rather than population-level incidence estimates. Results: Amongst 343 women (mean age 56.2 ± 7.4 years), nine cases (2.6%) of endometrial abnormalities were identified: four malignancies (1.2%), four endometrial hyperplasia without atypia (1.2%), and one complex atypical hyperplasia (0.3%). Only five cases (1.5%) required surgical intervention. All four endometrial cancers were Stage IA (FIGO 2009; three Grade 1, one Grade 2; no LVSI), corresponding to Stage IA2mNSMP under FIGO 2023. None required adjuvant therapy. Strikingly, 88.9% of all abnormal cases occurred within two years of HRT initiation, with no endometrial pathology identified amongst the 45 women using systemic HRT for more than five years—a temporal pattern not previously reported. Conclusions: In this retrospective all-age NHS cohort, the diagnostic yield of endometrial cancer was 1.2% in HRT users with unscheduled bleeding, with only 1.5% requiring surgical intervention. All cancers were early-stage (FIGO 2009 Stage IA; FIGO 2023 Stage IA2mNSMP) and required no adjuvant therapy. A previously unreported temporal clustering of pathology within the first two years of HRT initiation generates a hypothesis that early unscheduled bleeding may unmask pre-existing rather than HRT-induced endometrial abnormalities. These observations are hypothesis-generating and should not be interpreted as evidence of endometrial safety. These findings apply specifically to symptomatic women presenting for investigation and cannot be generalised to asymptomatic HRT users. Prospective validation in larger cohorts with baseline endometrial assessment is required before any clinical conclusions can be drawn. What This Study Adds: (1) A real-world cancer detection proportion of 1.2% in an all-age contemporary NHS cohort. (2) A previously undescribed temporal pattern with pathology clustering within two years of HRT initiation and no pathology in long-term users (n = 45), generating a testable hypothesis about pre-existing versus HRT-induced disease. (3) Dual FIGO 2009/2023 staging demonstrating that molecular classification added no treatment-discriminatory value in this early-detection context. Full article

The reliable identification of impairment relevant to safety-critical activities remains a significant challenge for public safety, motivating the exploration of unobtrusive and widely accessible sensing technologies. This study examines the viability of utilising inertial data acquired from consumer-grade smartphones to characterise gait disturbances associated with simulated visual impairment. The study simulates intoxication-related effects using alcohol-impairment goggles and does not involve the measurement of real alcohol intoxication. Two supervised experimental protocols were conducted in which participants traversed predefined walking routes under normal conditions and while wearing alcohol-impairment simulation goggles representing five manufacturer-declared blood alcohol concentration (BAC)-related goggle conditions plus a no-goggles control condition. An initial indoor trial, conducted in a structured corridor environment, yielded limited discrimination of gait dynamics due to strong spatial and visual stabilisation cues. To address this limitation, a subsequent outdoor experiment was conducted along a 100 m path lacking prominent visual reference points, resulting in motion patterns that more closely reflect unconstrained, real-world locomotion. Tri-axial accelerometer and gyroscope signals were recorded using smartphones, followed by artefact removal, segmentation, and standardisation to ensure inter-trial comparability. The resulting curated dataset comprises 290,919 multi-channel samples derived from 96 walking trials involving 16 participants and is released as an openly accessible resource to support further research in gait analysis and classification of gait alterations associated with simulated impairment. Model evaluation was performed using an 80/20 train–test split conducted within each traversal, with training and test windows originating from the same participant and walking session. Consequently, the reported results reflect within-subject performance instead of subject-independent generalisation. Multiple deep learning architectures combining convolutional feature extraction, bidirectional long short-term memory layers, and self-attention mechanisms were systematically evaluated. Using a subject-dependent evaluation protocol, the best-performing architecture achieved an accuracy of 71.4% and a weighted F1-score of 71.5% in distinguishing gait patterns associated with different levels of simulated visual impairment. The best-performing architectures yielded classification performance consistent with exploratory, low-stakes assessment of gait alterations associated with simulated visual impairment, using accelerometer data alone. These findings illustrate the feasibility of using smartphones as auxiliary tools for exploratory, low-stakes screening or educational applications and contribute a publicly released dataset and benchmark results to facilitate methodological advancement in inertial sensor-based gait impairment analysis. Full article

High-Andean grasslands in the Central Andes of Peru are severely degraded by Lepidium meyenii (maca) cultivation, compromising pasture structure and forage availability for sustainable livestock production. A factorial field experiment evaluated restoration timing and pasture-oriented seed mixtures by manipulating resting time after abandonment (0, 1, 2, and 3 years) and restoration treatment (control; Festuca dolichophylla monoculture; full mixture of Dactylis glomerata + Lolium spp. + Trifolium repens + F. dolichophylla; and mixture without F. dolichophylla) across 64 plots. Vegetation was assessed eight months after seeding, and responses were analysed with ordination, PERMANOVA with restricted permutations, PERMDISP, and generalised linear models and mixed-effects models for diversity metrics. Community composition differed significantly among resting times and seed treatments, with resting time explaining the largest proportion of variance (R² = 0.353), followed by treatment (R² = 0.236), while the interaction was significant but smaller (R² = 0.102, p = 0.002). PERMDISP detected significant differences in multivariate dispersion for both Resting Time and Treatment, indicating that compositional differences may reflect both centroid shifts and heterogeneity among groups. Passive recovery and Festuca-only plots showed slower, more variable compositional change, whereas productive mixtures produced clearer, treatment-specific trajectories over time, suggesting possible divergence in community development patterns, rather than providing formal evidence of distinct alternative stable states. Establishment was consistently high for D. glomerata and Lolium spp., supporting rapid ground cover, which may be associated with short-term forage potential, while F. dolichophylla showed chronically low establishment consistent with limited germination performance. The invasive Pennisetum clandestinum was most pronounced under passive recovery and was reduced under seeded mixtures, suggesting a potential competitive suppression effect. Overall, early seeding with productive mixtures appeared to influence community assembly trajectories, while resting time remained the dominant driver of compositional variation, suggesting potential implications for restoration management in maca-degraded landscapes, although outcomes related to sustainable grazing systems were not directly evaluated. Full article

Recent research in grounded language learning has seen remarkable success due to advances in large vision and language models (VLMs). However, these models (i) are extremely costly to train and update; (ii) struggle with generalisation; and (iii) do not support continual learning. In this paper, we introduce baby-ds integrating the Dynamic Syntax (DS) framework with automated planning within the multimodal BabyAI platform as a testbed. We provide methods whereby DS lexicons are induced continually from teacher demonstrations within BabyAI. We study (i–iii) by experimenting with the compositional complexity of natural language instructions in the data to compare data efficiency, generalisation, and continual learning properties of baby-ds with a simple neural model. The results show that the baby-ds model: (i) needs much less data than the neural model to reach threshold performance; (ii) generalises much faster to more complex instructions; and (iii) is a more effective continual learner. We argue that it is the attendant linguistic bias within DS and the rich inferential power of TTR that enables (i–iii), highlighting the importance of further research on hybrid grammar–neural approaches. Finally, we discuss several important limitations of baby-ds and sketch a path forward for further DS research. Full article