Abstract: Large language models (LLMs) have advanced medical reasoning, but static question-answering performance remains insufficient for clinical workflows that require evolving patient-state tracking, evidence integration, role coordination, and accountable decisions. LLM-based medical multi-agent systems (MAS) are being developed to move AI from isolated answer generation toward workflow-level clinical intelligence by combining role specialization, memory, tool use, retrieval, communication, and orchestration. This Review maps LLM-based medical MAS across diagnosis, treatment decision support, imaging, monitoring, surgery, hospital workflow automation, evidence synthesis, medical education, and safety governance. We further synthesize key architectures for collaboration, knowledge-augmented evidence chains, multimodal integration, privacy-preserving coordination, and adaptive optimization, together with evaluation strategies spanning outcomes, process quality, robustness, efficiency, human comparison, and temporal backtesting. We argue that medical MAS should be evaluated not as larger LLM workflows, but as clinical coordination infrastructures that redistribute evidence, responsibility, and risk across human-AI teams. Their value depends on auditable evidence chains, controllable orchestration, explicit role accountability, and clinician oversight, rather than autonomous answer generation. Before routine clinical use, future work should prioritize traceable evidence chains, human oversight, privacy-preserving collaboration, standardized reporting, regulatory readiness, and prospective clinical validation.

Abstract: Oncogenic viruses contribute to approximately 15–20% of human cancers globally, with their impact falling disproportionately on populations in Sub-Saharan Africa. In this region, cervical cancer, hepatocellular carcinoma, endemic Burkitt lymphoma, and Kaposi sarcoma represent major causes of cancer-related morbidity and mortality, driven by persistent infection with human papillomavirus (HPV), hepatitis B and C viruses (HBV/HCV), Epstein–Barr virus (EBV), Kaposi sarcoma–associated herpesvirus (KSHV), and human T-lymphotropic virus-1 (HTLV-1). This review synthesizes current insights into the immunological mechanisms that underpin viral carcinogenesis in Africa, emphasizing how defective viral clearance, chronic immune activation, and immune evasion arise from the convergence of region-specific co-infections, host genetic diversity, and environmental exposures. We examine the mechanistic roles of HIV-associated CD4⁺ T cell depletion, malaria-induced perturbation of antiviral T cell immunity, helminth-driven T helper 2 polarization, and tuberculosis-associated inflammatory signaling in promoting viral persistence and malignant transformation. In addition, the influence of the extensive diversity of African human leukocyte antigens (HLA) and cytokine gene polymorphisms on antiviral immune responses and cancer susceptibility was discussed. We also assessed how virus-associated tumors establish profoundly immunosuppressive microenvironments characterized by impaired antigen presentation and the dominance of immune checkpoint pathways. Finally, we examined how gaps in vaccination, screening, and diagnostic capacity intersect with immunological vulnerability across Africa, contributing to the burden of infection-associated cancers. These challenges position Africa as a critical setting for developing targeted, genotype-inclusive public health interventions and reducing global cancer disparities through advances in immunoprevention and immunotherapy.

Abstract: Air pollution is a public health threat that requires urgent action. This study conducted a bibliometric analysis of air pollution and human health in Indian cities to examine trends and the geography of the scientific literature, the evolution of research, and co-occurrence patterns of pollution sources, types, and health impacts. Furthermore, a narrative review of air pollution mitigation strategies was conducted using scholarly articles, policy documents, and reports. Relevant publications from the Web of Science (WoS) and Scopus databases were downloaded. The search identified 3307 articles published between 1987 and 2024, of which 172 met the inclusion criteria. The bibliometric analysis was conducted using VOSviewer and R software. The results indicate a steady rise in studies on air pollution and health issues in India. Initially, research concentrated on various sources and types of pollution, subsequently transitioning to the evaluation of exposure risks, risk assessment, and health implications, ultimately narrowing its focus to risk assessment concerning human health. Over the course of forty years, there has been a growing emphasis on the influence of indoor air quality, including ‘PM2.5’, ‘PM10’, dust, chemical pollutants, heavy metals, and exhaust dust, on human health. Research on pollution-related health effects has moved from examining general impacts to focusing on long-term, chronic consequences of pollutant exposure. Notably, most studies are centred in large metropolitan areas, whereas medium and small towns are underrepresented. Urban areas face severe air-quality challenges, requiring strategies such as monitoring pollution, promoting renewable energy, reusing materials, installing green walls or buffers in pollution zones, improving transport infrastructure, and reducing dust with grass covers. This study underscores the importance of implementing effective air pollution control measures across various geographic regions and integrating air pollution mitigation strategies into comprehensive urban development and planning frameworks.

Abstract: Acute coronary syndromes (ACS) remain a major cause of cardiovascular mortality and long-term morbidity worldwide despite substantial advances in revascularization techniques, antithrombotic therapies, and lipid-lowering strategies. Although contemporary management has significantly improved survival, a considerable residual risk of recurrent ischemic events, adverse ventricular remodelling, heart failure, and cardiovascular death persists following ACS. Consequently, accurate risk stratification has become increasingly important to guide individualized secondary prevention and optimize long-term outcomes. Traditional risk assessment after ACS relies largely on clinical characteristics, conventional cardiovascular risk factors, and established risk scores such as GRACE and TIMI. However, growing evidence indicates that several emerging biomarkers and pathophysiological pathways provide additional prognostic information beyond conventional models. These markers reflect distinct biological processes including residual inflammatory activity, myocardial injury and remodelling, renal dysfunction, metabolic impairment, and visceral adiposity. This review summarizes current evidence regarding both established and emerging prognostic markers after ACS across six major pathophysiological domains: lipid abnormalities and traditional cardiovascular risk factors; systemic inflammation and residual inflammatory risk; myocardial injury and adverse ventricular remodelling; cardiorenal dysfunction; visceral adiposity; and anti-inflammatory therapeutic strategies. For each domain, we discuss underlying biological mechanisms, key findings from major clinical trials and registries, and the potential incremental value of biomarkers in contemporary risk stratification. Collectively, these markers support a multidimensional approach to post-ACS risk assessment and may contribute to more personalized therapeutic strategies aimed at reducing residual cardiovascular risk.

Abstract: Fracture and crack propagation in flexible shells under extreme loading represent a fundamental challenge in continuum mechanics. Traditional shell fracture theories rely heavily on local coordinate systems and asymptotic expansions, often entangled in the contradiction between three-dimensional solid fracture and two-dimensional shell theory. Taking the geometrically exact Kirchhoff-Love shell theory based on fiber bundles and differential forms previously established by the author as a starting point, this paper strictly generalizes it to a two-dimensional mid-surface manifold topology containing evolving cracks. We model through-cracks as evolving internal boundaries and one-dimensional submanifolds on the two-dimensional mid-surface manifold, introducing a rigorous kinematic mapping for crack propagation. In terms of dynamics, based on the elastic strain energy on the two-dimensional mid-surface, we derive a geometrically exact two-dimensional Eshelby configuration stress tensor and express it as a vector-valued configuration stress 1-form. Through the generalized virtual work principle applied to the variation of the crack front, a coordinate-independent J-integral (energy release rate) is naturally defined. Based on the Griffith criterion and the maximum energy release rate principle, this paper strictly derives the control equations for the crack propagation direction vector and propagation velocity on the tangent space of the manifold. This theory implicitly contains the complex curvature-fracture coupling within the structures of exterior differentiation and pullback metrics. Furthermore, we present a complete Discrete Exterior Calculus (DEC) discretization framework for the theory.

Abstract: Background The 2026 Bundibugyo virus disease (BVD) outbreak, caused by Bundibugyo virus (BDBV), was declared by the Democratic Republic of the Congo (DRC) Ministry of Public Health on 15 May 2026 and subsequently reported across multiple health zones in Ituri, Nord-Kivu, and Sud-Kivu provinces. The World Health Organization subsequently determined that the event constituted a public health emergency of international concern. Outbreak situation reports generate large volumes of health-zone-level data, but operational teams require a standardized mechanism for translating those data into prioritized, role-specific field guidance.Methods This study developed the Bundibugyo Virus Disease Operational Intelligence System (BVD-OIS), a three-module framework built on an eight-domain Health Zone Operational Priority Index (HZ-OPI, version 1.1). The first additional module, the Field Action Prioritization Tool (FAPT), adds a trajectory layer that classifies changes in a health zone’s composite score between reporting cycles and a binding constraint engine that identifies the response domain most likely to limit outbreak control. The second additional module, the Response Intelligence Audit (RIA), flags health-zone domains where available data are insufficient to support confident scoring. All modules use aggregate, health-zone-level indicators from public reports and do not use individual-level data.Results The framework produces two standardized outputs: a one-page Health Zone Action Card combining priority tier, trajectory, binding constraint, data completeness, and time-stratified role-specific actions; and a Multi-Zone Priority Summary ranking scored health zones for coordination settings. This manuscript specifies the domain scoring rubric, priority tiers, 21-day inactivity override, trajectory classification scheme, binding constraint hierarchy, action translation matrix, data gap detection logic, and a fully worked illustrative example using hypothetical health-zone data.Conclusions The BVD-OIS provides a reproducible pathway from routinely reported outbreak surveillance indicators to field-level operational guidance. The framework is descriptive and not a validated predictive model. It is intended for transparent operational learning, local adaptation, and prospective validation during filovirus and other high-consequence outbreak responses.

Abstract: This study examines social media communication in Kenya to assess how dryland pastoral concerns are represented in national climate policy dialogues. Approximately 80% of Kenya’s land area consists of drylands that support pastoral livelihoods, yet recurring droughts and increasingly variable rainfall threaten these social‐ecological systems. Although Sustainable Development Goal 13 (Climate Action) emphasizes adaptation, progress is typically measured using national indicators that may overlook local priorities. We analyzed digital communications from pastoral nongovernmental organizations, civil society organizations, newspapers, and government agencies using an adaptation pathways framework to examine links among climate risks, governance, and decision‐making. Five themes emerged: water security, rangeland degradation and mobility, climate communication, gender and social inclusion, and climate finance. Kenya has developed relatively strong national climate policies, but social media reveals a persistent implementation gap between national policy and local pastoral priorities. Pastoral communities consistently emphasize Indigenous and local knowledge, women’s leadership, community participation, and locally appropriate adaptation strategies. Social media provides a novel source of evidence for identifying these local priorities and informing more participatory, place‐based climate adaptation.

Abstract: The discovery of the glymphatic system and meningeal lymphatic vessels established that the brain possesses an organized lymphatic clearance network that drains to the deep cervical lymphatics, and that disruption of this system contributes to the accumulation of pathogenic proteins implicated in neurodegeneration. Cervical lymphatic reconstruction has shown early neurological benefit in Alzheimer’s and Parkinson’s disease, and preclinical work demonstrates that noninvasive manipulation of superficial cervical lymphatics increases cerebrospinal fluid outflow. Whether noninvasive cervicofacial lymphedema therapy can produce comparable neurological effects remains unknown. We report two men with Parkinson’s disease (Hoehn and Yahr stages 2.5 and 3) treated with a standardized cervicofacial lymphedema therapy protocol targeting cervical and facial lymphatic pathways. Total MDS-UPDRS scores improved in both patients, by 19 points (−34.5%) in Patient 1 and 5 points in Patient 2, each meeting or exceeding the minimal clinically important difference. Improvements were reproducible and session-linked, and were accompanied by motor and non-motor gains in cognition, alertness, mood, energy, and sleep; interruption of therapy was associated with return toward baseline. Cervicofacial lymphedema therapy may represent a low-risk, noninvasive strategy for modulating brain lymphatic clearance in Parkinson’s disease.

Abstract: Twin Transition denotes the dual transformation toward digitalization and sustainability. The concept is promising but lacks empirical evidence whether digital infrastructure really improves environmental conditions or on the contrary. This study adopts the difference-in-difference (DID) model to examine the impact of China’s policy shock of 5G Pilot Program on enterprises’ environmental, social, and governance (ESG) performance. The policy was carried out in batches across 40 scientifically selected cities, forming quasi-natural experiment. Using panel data covering 50,838 A-share listed firms from 2009 to 2024, we find adopting 5G improved the ESG performance of enterprises. Mechanism tests show two mediating paths: accelerated corporate digital transformation and promoted green technological innovation. Heterogeneity analysis showed stronger effect for non-state-owned enterprises, in high-tech industries, and in the eastern region of China. This study provides empirical evidence at firm level for the interaction mechanism between digitalization and sustainability, lending reference for informed strategy formulation and governance.

Abstract: Household food waste remains a huge challenge for solid waste management in municipalities worldwide, especially in the Global South. Existing studies that measured food waste (FW) in cities are scarce, have limited geographic scope, and have limited timeframes. In that direction, the current investigation provides data on the FW composition of nine regions of the municipality of Rio de Janeiro (Brazil), based on a three-year sampling across 155 neighborhoods. Waste samples were collected from 2021 to 2023. In total, about 24,038 kg (fresh weight) were analyzed. Results showed that FW accounts for an average of 47.7±1.9% of household waste in the study period. The FW composition in the city of Rio de Janeiro ranged from 60.3 – 76.5% for fruits, vegetables, and salads, 15.0 – 25.1% for fine aggregate (small-sized food residues < 2.54 cm, like rice, beans, grains, and fragmented food particles), and 3.2 – 5.8% for proteins (discarded animal-based protein foods like chicken and meat). The chi-square good-ness-of-fit test was applied to evaluate whether the FW composition in each of the nine regions differed from the mean FW composition of the Rio de Janeiro municipality. The findings revealed statistically significant differences (p-value < 0.05) in the average FW fractions in specific regions and years compared with the city’s average composition. Thus, one of the key takeaways of this investigation was that the percentages of discharged food waste fractions vary over time and across locations, even within the same municipality. The present research took a first step toward understanding the food waste problem in Rio de Janeiro (Brazil) and underscores the importance of monitoring food waste data to guide the development of locally specific strategies for sustainable urban food systems, including waste prevention, recycling, and food recovery.

Abstract: Cryptosporidium spp. are important zoonotic parasite causing gastrointestinal diseases in humans and animals. Cattle are major hosts for Cryptosporidium species such as C. parvum, C. bovis, C. ryanae, and C. andersoni, with C. parvum being the most pathogenic and a key zoonotic agent. However, systematic epidemiological data on Cryptosporidium in dairy cattle from Anhui Province remain scarce. This study aims to investigate its prevalence and genetic diversity to support effective control strategies. Cryptosporidium species and subtypes were identified using the small subunit ribosomal RNA (SSU rRNA) and 60 kDa glycoprotein (gp60) genes. PCR testing revealed an overall Cryptosporidium infection rate of 35.4% (338/956) across four dairy farms in Anhui Province, with significant regional variation (p < 0.01). Bengbu exhibited the highest infection rate (61.5%), while Fuyang showed the lowest (8.6%). Age-specific distribution indicated that calves aged 2-6 months had the highest infection rate (50.2%). Four Cryptosporidium spp. were identified: C. parvum, C. andersoni, C. ryanae, and C. bovis, with C. parvum and C. bovis being predominant. The IId subtype family dominated C. parvum isolates. Seven gp60 subtypes were detected in C. ryanae, and five distinct subtypes were identified in C. bovis. The study revealed distinct geographical and age-associated variations in bovine cryptosporidiosis in Anhui Province, highlighting the need for targeted control strategies in high-risk regions and young calves.

Abstract: Backgroung / Objective: Radiomics offers a powerful, non-invasive approach for extracting quantitative features to predict lesion phenotypes. This study performs a radiomic characterization of breast tissue using synchrotron radiation breast computed tomography (SR-bCT). Method: Four mastectomy samples were imaged at the Australian Synchrotron (ANSTO) using five energies (25, 32, 35, 40, and 60 keV). From 10 slices per energy for each sample, 1546 regions of interest (ROIs) were extracted across four tissue subtypes: fatty, glandular, fibrous, and microcalcified. Using Pyradiomics, 93 features were initially calculated and then reduced to 34, eliminating highly correlated variables to reduce redundancies. Seven linear regression models and a discriminant analysis evaluated subtype tissue separation and radiomic characterization across individuals and combined samples and energies with a good fit (0.81≤ R ≤ 0.98). Results: The study identified 6 robust possible biomarkers independent of sample variability and energy levels, whose mean values are significantly different among the 4 tissue subtypes (clusters, p< 0.001). The selected biomarkers were: 90th Percentile, Kurtosis, Skewness, GLCM IDM, GLDM LDE, and NGTDM Coarseness. These metrics successfully differentiated all tissue pairs (microcalcifications/fat, microcalcifications/gland, microcalcifications/fiber, fat/gland, fat/fiber and gland/fiber), with p < 0.05 inside the most general regression model and p< 0.0001 with the linear discriminant separation in clusters. Conclusion: Findings indicate that some first-order and second-order texture metrics reflecting global dependencies remain stable across experimental conditions. Conversely, fine-texture metrics are highly sensitive to sample energy changes, limiting their generalizability. These results align with successful biomarkers in mammography and validate the potential of radiomics in SR-bCT characterization.

Abstract: Plants of the genus Crataegus have been used in traditional medicine to treat different health conditions, mainly cardiovascular diseases. Standardized extracts from this genus are marketed in Europe and Asia for the treatment of heart failure. In recent years, our research group has demonstrated that Crataegus gracilior, C. rosei, and C. mexicana exert significant vasorelaxing effects and that their most abundant and vasorelaxant constituents are the triterpenic acids they contain. Therefore, an HPLC-DAD analytical method was developed to simultaneously identify and quantify euscaphic, maslinic, corosolic, oleanolic, and ursolic acids, the main chemical constituents of the leaves of these three Mexican Crataegus species. Euscaphic acid was found to be the main compound in both C. rosei and C. mexicana while ursolic acid in C. gracilior. Therefore, these two acids were selected as their most suitable chemical and pharmacological markers. Accordingly, the developed method was validated for the two acids following the ICH Q2(R1) and USP guidelines. This method can be used for quality control of any crude commercial drugs produced from these species in the future.

Abstract: The study introduces a solar-powered electric cooking system designed for households in sub-Saharan Africa. It uses a 330W photovoltaic (PV) panel directly connected without a battery, to ten parallel Positive Temperature Coefficient (PTC) heating elements mounted on a 20 cm diameter circular aluminium plate to form a “PTC hot plate.” An Arduino controller manages the load and records system data (current, voltage, power, and temperature). Cooking tests under different weather conditions showed practical performance: 0.5 kg of beans cooked in ~3 h using 1.4 kWh; 1.0 kg of beef in 1.5 h using 1.1 kWh; and 0.5 kg of rice in 1 h using 0.73 kWh. Boiling 2 liters of water took about 25 min. Sequential cooking (rice then beans), starting at 9:45 h was completed by 14:30 h. The average cooking efficiency was evaluated to be about 54%, dependent on the duration of cooking and the food type being cooked. A minimum solar irradiance of about 400 W/m2 per day was required for effective cooking. These results demonstrate that the solar PV-PTC cook stove is a viable and promising solution for meeting household cooking needs in the sub-Saharan African region.

Abstract: Background/Objectives: Immunisation coverage in the Democratic Republic of the Congo (DRC) consistently falls short of global standards, placing the country among those with the highest prevalence of unvaccinated or “zero-dose” children. This study aimed to generate high-resolution (1 km × 1 km) geospatial estimates of coverage for the first to third doses of diphtheria-tetanus-pertussis vaccine (DTP1-3) and the first dose of measles-containing vaccine (MCV1), along with estimates of zero-dose and under-vaccinated children to support vaccination programming in DRC. Methods: We used Bayesian geostatistical modelling to integrate data from the 2023 Enquête de Couverture Vaccinale (ECV) survey with geospatial covariates and harmonized population datasets. The model generated vaccine coverage and dropout rates at 1×1 km resolution while accounting for spatial dependence and prediction uncertainty. Grid-level estimates were aggregated to health areas, health zones, and provinces using population-weighted averages, and combined with under-one population estimates to quantify zero-dose and DTP-under-vaccinated children. Results: Grid-level coverage estimates revealed substantial geographic heterogeneities across the DRC. 56.4% of mapped health areas achieved DTP1 coverage above 80%, but this declined to 32% for DTP2, 15% for DTP3, and 8.3% for MCV1. Around 60.5% of the health zones achieved DTP1 coverage≥80%, while only 2.7% fell below 40%, indicating that substantial disparities persist across these zones. Nationally, based on constrained coverage estimates (only areas with identified built settlements) applied to an estimated 4.2 million children under one year in 2024, 18.4% had not received DTP1, 20.7% were DTP-under-vaccinated (received DTP1 but not DTP3), and 46.8% remained unvaccinated for MCV1. Conclusions: The prevalence of large numbers of zero-dose children, together with increased dropout rates, underscores systemic issues in access and follow-up initiatives. These findings are critical for microplanning and targeted outreach to achieve equitable immunisation coverage in line with the Immunisation Agenda 2030’s goal of leaving no child behind.

Abstract: The paper presents the investigations results of the transport process of the fluid flow through a pipeline under conditions of temperature gradient between the internal and external environments and continuous gas generation at the contact boundary of the transported media. In the non-isothermal flow case, a slippage effect will impact the flow velocity, the pressure and the temperature distributions in variable cross section pipes. Laboratory experiments were conducted in order to study the effects of the gas nucleus at the pipe walls on the hydrodynamic characteristics of the fluid flow. It is shown that a throughput capacity of the pipe is affected by the temperature difference between the oil and the pipe walls. At certain temperature gradient on the border layer, the pipe capacity reached a maximum value.

Abstract: Adaptive bitrate streaming over HTTP (DASH) requires content-aware bitrate ladders to balance bandwidth utilisation and quality of experience, particularly for computationally demanding codecs such as H.266/VVC. This paper introduces MHCA-EGB (Multi-Head Cross-Attention with Ensemble Gradient Boosting), an integrated framework that systematically combines established attention-based fusion, multi-scale pooling, and ensemble classification techniques in a purpose-designed architecture that predicts Pareto-optimal bitrate resolution pairs for H.266-encoded video delivered over DASH by jointly modelling video content complexity and compression-domain artefacts. The proposed architecture extends the dual-path 2D–3D CNN paradigm with three key contributions: (i) a Multi-Head Cross-Attention fusion module that replaces naïve channel concatenation, enabling learned bidirectional feature alignment and content-adaptive emphasis of quality-discriminative representations; (ii) a Temporal Pyramid Pooling layer that captures multi-scale temporal dynamics from short-burst motion to long-range scene transitions at 2-frame, 4-frame, and 8-frame granularities; and (iii) a stacked ensemble classifier combining XGBoost, LightGBM, and CatBoost with a logistic regression meta-learner for robust bitrate cluster assignment. Comprehensive evaluation on 101 diverse 4K UHD sequences from four benchmark datasets, encoded with VTM 22.0 at seven resolutions and eleven QP values (7,777 total encodes), demonstrates that MHCA-EGB achieves an average BD-Rate of −5.47% relative to the exhaustive convex hull (a value attributable to the BD-Rate polynomial fitting methodology when the predicted and reference ladders use overlapping operating-point subsets; all predicted points are physically encoded and verified) while reducing encoding time by 98.7% with only 0.12 VMAF regret (fifty times below the just-noticeable difference threshold). Ablation analysis confirms that cross-attention fusion contributes the largest novel gain (+1.42% BD-Rate), followed by temporal pyramid pooling (+0.79%) and ensemble stacking (+0.55%), with a Pearson correlation of ρ=0.87 between content complexity and BD-Rate magnitude confirming that the framework delivers the greatest value on high-complexity premium content.

Abstract: Type 2 diabetes mellitus (T2DM) is a priority public health issue that requires accurate, efficient, and interpretable predictive tools to support the early identification of at-risk cases. Nevertheless, the comparison of machine learning models in this domain is frequently constrained by heterogeneous protocols, a reliance on overall accuracy, and an inadequate integration of metrics for agreement, calibration, robustness, and interpretability. This study employed a comparative approach to evaluate 10 ensemble and boosting models for the binary classification of T2DM under a unified experimental protocol. The evaluation employed nested cross-validation with 10 external folds and 3 internal folds, a 20% internal holdout, and complementary metrics of performance, discrimination, agreement, calibration, and computational cost. In the context of nested cross-validation, the Random Forest algorithm demonstrated a superior performance, attaining the highest average weighted F1-score (92.51 ± 5.80%), the highest Matthews' correlation coefficient (MCC) (0.840 ± 0.120%), and high discriminative power (weighted area under the receiver operating characteristic curve (ROC-AUC) = 97.46 ± 2.12%). Consequently, it was selected based on a predefined composite criterion. CatBoost achieved the highest weighted ROC-AUC (97.91 ± 2.39%) and area under the precision-recall curve (PR-AUC) (96.60 ± 3.61%), while Extra Trees demonstrated performance that was virtually equivalent to that of the selected model. The Friedman test revealed significant overall differences among models (χ² = 27.898607; p = 0.000992). However, the Nemenyi post-hoc test indicated that the leading models were statistically comparable to one another and that the significant differences were concentrated relative to AdaBoost. In the holdout set, the calibrated Random Forest attained an accuracy of 0.882, a balanced accuracy of 0.880, a weighted F1-score of 0.884, an ROC-AUC of 0.943, a PR-AUC of 0.879, a κ of 0.736, an MCC of 0.739, a Brier score of 0.091, and an expected calibration error of 0.130, with a recall of 0.88 for the positive class. Explainability analyses employing SHapley Additive Explanations (SHAP) and Local Interpretable Model-Agnostic Explanations (LIME) techniques have been demonstrated to offer traceability and predictive plausibility, whilst eschewing the assumption of causality. The findings of this study suggest that Random Forest, Extra Trees, and CatBoost represent robust alternatives for the tabular classification of T2DM. However, it is important to note that their clinical application requires multicenter external validation, local recalibration, and prospective evaluation.

Abstract: Spatial self-organized patterns are ubiquitous features of vegetation ecosystems, and cyclic nontransitive competition serves as a crucial intrinsic mechanism for sustaining biodiversity. However, existing studies lack cross-scale comparative analyses of vegetation territorial competition based on multiple models. This study combines lattice models and continuous differential equation models to investigate the territorial occupation dynamics and spatial evolution of vegetation communities driven by cyclic competition. The results demonstrate that cyclic competition acts as a core mechanism maintaining vegetation biodiversity, which enables the self-organization of stable spiral wave patterns in space and supports the long-term dynamic coexistence of multiple species. Discrete and continuous models exhibit highly consistent macroscopic dynamical behaviors, which reveal the intrinsic dynamical characteristics of cyclic competitive systems. By integrating microscopic lattice simulation and macroscopic differential equation analysis, this study verifies that spiral waves represent a highly robust species coexistence mode and clarifies the coupled regulatory effects of species richness and stochasticity on system evolution. The findings further deepen the understanding of the complexity of vegetation ecosystems and provide important theoretical references for subsequent theoretical derivation and field observational research on vegetation community competition and evolution.

Abstract: This study presents a 3D geological model and structural interpretation of the Masa Valverde volcanogenic massive sulphide (VMS) deposit in the Iberian Pyrite Belt. The deposit is hosted by felsic porphyritic volcanic rocks, volcanic tuffs and black shales. A 3D geological model of the orebodies and host rocks, constructed from 145 drillcore logs, allowed us to build 16 cross-sections spaced 100 m apart, and constrain the mineralisation geometry and its structural evolution. Mineralization formed during Early Carboniferous transtensional tectonics within an extensional basin, where an extensional duplex structure controlled the development of the primary massive sulfide body and its associated stockwork. Subsequent counterclockwise rotation of the principal stress axes reactivated extensional faults as reverse faults during tectonic inversion. This deformation strongly modified the VMS system through buttressing, generating extensive open spaces and promoting brecciation and recrystallization of both the stockwork and massive sulfides. These processes produced a new paragenesis dominated by chalcopyrite and sphalerite, with minor galena among other minerals, which cemented the breccias, partially replaced earlier mineral assemblages, and filled open fractures. The resulting Cu-Zn enrichment, spatially associated with buttressed zones, provides new insights into ore remobilization with direct implications for the development of the ongoing underground mine.