Search Results (157)

This study addresses the challenge of accurate classification under missing data conditions by integrating multiple imputation strategies with discriminant analysis frameworks. The proposed approach evaluates six imputation methods (Mean, Regression, KNN, Random Forest, Bagged Trees, MissRanger) across several discriminant techniques. Simulation scenarios varied in sample size, predictor dimensionality, and correlation structure, while the real-world application employed the Cirrhosis Prediction Dataset. The results consistently demonstrate that ensemble-based imputations, particularly regression, KNN, and MissRanger, outperform simpler approaches by preserving multivariate structure, especially in high-dimensional and highly correlated settings. MissRanger yielded the highest classification accuracy across most discriminant analysis methods in both simulated and real data, with performance gains most pronounced when combined with flexible or regularized classifiers. Regression imputation showed notable improvements under low correlation, aligning with the theoretical benefits of shrinkage-based covariance estimation. Across all methods, larger sample sizes and high correlation enhanced classification accuracy by improving parameter stability and imputation precision. Full article

Fire has had a profound impact on Australia’s landscapes and biodiversity since the late Tertiary. Indigenous (Aboriginal) people have lived in Australia for at least 65,000 years and fire is an integral part of their culture and cosmology. In 2015, an Indigenous cultural burn was undertaken by Banbai rangers at Wattleridge Indigenous Protected Area, New England Tablelands, NSW. We compared the impact of this burn on the composition, cover, abundance, and species richness of dry sclerophyll vegetation and fuel hazard, with a hazard reduction burn at nearby Warra National Park, using a Before-After-Control-Impact experimental design. Our study found that the low-severity cultural burn and moderate-severity hazard reduction burn reduced fuel loads but did not have a significant impact on the composition of the vegetation overall or the herb layer. The hazard reduction burn had a significant impact on shrub and juvenile tree (woody species) cover, while the abundance of woody species was significantly affected by both fires, with a mass germination of ‘seeder’ species, particularly after the cultural burn. The long unburnt fire regime at Wattleridge may have made the vegetation more responsive to fire than the more frequently burnt vegetation at Warra, through accumulation of seed in the seed bank, so that the patchy cultural burn had a greater impact on woody species abundance. In terms of ecological and bushfire management outcomes, this study provides evidence to support claims that Indigenous cultural burning decreases fuel loads, stimulates regeneration of shrubs and trees, and manages at a local, place-based scale. We recommend cultural burning as a key management tool across Indigenous Protected Areas and other land tenures, with its implementation monitored and adaptively managed through two-way science, to foster fire regimes that are both culturally and ecologically beneficial. This is a vital element of our resilience in the Pyrocene and a significant step toward decolonizing science and land management. Full article

Introduction: Overuse injuries are a growing concern among adolescent soccer players, with the repetitive nature of the sport placing significant physical demands on young athletes. These injuries can have long-term implications for physical development, performance, and overall well-being. This narrative synthesis aimed to evaluate the existing literature on the epidemiology, risk factors, and management strategies for overuse injuries in adolescent soccer players. Methods: A comprehensive literature search was conducted using PubMed and Embase. A total of 123 articles were identified, 27 of which met the inclusion criteria after screening. Studies focusing on overuse injuries in adolescent soccer players aged 10–18 years were included, while those addressing acute injuries, non-soccer populations, or adult athletes were excluded. Relevant quantitative and qualitative data were extracted and evaluated. Due to heterogeneity in study designs and outcomes, findings were narratively synthesized rather than meta-analyzed. Results: The period around peak height velocity (PHV: 11.5 years in girls, 13.5 years in boys) was consistently identified as a high-risk window, with seven studies demonstrating a significantly increased incidence of overuse injuries. Additional risk factors included leg length asymmetry, truncal weakness, early sport specialization, high ratios of organized-to-free play, and increased body size. Injury burden was greatest for hamstring and groin injuries, often leading to prolonged time lost from play. Preventive interventions such as plyometric training, trunk stabilization, and structured load monitoring demonstrated reductions in injury incidence in several prospective studies, though protocols varied widely. Conclusion: This narrative synthesis highlights PHV as the most consistent risk factor for overuse injuries in adolescent soccer players, alongside modifiable contributors such as training load, sport specialization, and free play balance. Evidence supports neuromuscular training and structured monitoring as promising preventive strategies, but there remains a lack of standardized, evidence-based protocols. Future research should focus on optimizing and validating interventions, integrating growth and load monitoring, and leveraging emerging approaches such as machine learning-based risk prediction. Full article

The periphery of Lomami National Park in the Democratic Republic of the Congo (DR Congo) is experiencing intense and increasing hunting pressure, driven by both local subsistence needs and growing urban demand for bushmeat. This situation poses a serious challenge to sustainable natural resource management and underscores the need to realign protected-area policies with the realities faced by surrounding communities. In the absence of comprehensive ecological monitoring, this study used hunters’ perceptions to assess the current availability of mammalian wildlife around the park. From October to December 2023, surveys were conducted using a snowball sampling method with 60 hunters from nine villages bordering the park. Results show that hunting is a male-dominated activity, mainly practiced by individuals aged 30–40 years, with firearms as the primary tools. It occurs both in the park’s buffer zones and, alarmingly, within its core protected area. This practice has contributed to the local disappearance of key species such as African forest elephant (Loxodonta cyclotis), African buffalo (Syncerus caffer), and African leopard (Panthera pardus pardus), and to the marked decline of several Cephalophus species. These patterns of overexploitation reveal critical weaknesses in current conservation strategies and point to the urgent need for integrated, community-based resource management approaches. Strengthening law enforcement, improving ranger support, and enhancing participatory governance mechanisms are essential. Equally important is the promotion of sustainable alternative livelihoods—including livestock farming, aquaculture, and agroforestry—to reduce hunting dependence and build long-term resilience for both biodiversity and local communities. Full article

The significant impact of road traffic accidents on public health requires clear and effective policies to combat them. However, public action can only be truly effective when supported by robust monitoring tools. This project aims to evaluate the effectiveness of a set of machine learning algorithms in predicting road accidents in Switzerland, utilizing open-access Confederation drive crash databases combined with environmental and socio-economic factors. Three different algorithms are tested: Logistic Regression Model (LRM), Random Forest with Ranger (RF), and Artificial Neural Network (ANN) with Keras. Among the predictive factors, road types are shown to be of high importance in all models. Regarding model performance, all the applied algorithms show a high level of accuracy, with all models achieving over 90%. The Random Forest algorithm, optimised using the Ranger application, exhibited the best performance, particularly in terms of specificity (0.88 compared to 0.34 and 0.40 for LRM and Keras, respectively) and negative predictive value (0.96 compared to 0.65 for LRM and 0.68 for Keras). These results suggest that this approach could support public policy for traffic management, if data collection and sharing activities are constantly carried out. Full article

Background: Peripheral nerve injuries, especially involving the facial nerve, present unique reconstructive challenges due to their complex functional demands and limited regenerative potential. While autografts remain the gold standard, their drawbacks—such as donor-site morbidity and limited availability—have driven interest in processed nerve allografts. Acellular grafts, in particular, offer promising off-the-shelf alternatives without the need for immunosuppression. Methods: We conducted a narrative review of the literature (1990–2023), identifying 55 peer-reviewed studies via PubMed, Embase, and Cochrane Library. The studies included clinical and preclinical work on motor nerve regeneration using processed nerve allografts, with particular attention to outcomes in facial nerve repair. Two independent reviewers conducted abstract screening, full-text review, and data extraction. Results: Processed nerve allografts show encouraging motor recovery in gaps under 50 mm, with recovery rates of up to 85% reported. Outcomes decrease significantly in longer gaps (>50–60 mm) and in complex cases, including facial nerve repairs, where evidence remains sparse and largely extrapolated from broader motor nerve data. Registry data (e.g., RANGER) support their use but are limited by heterogeneity and lack of randomization. Conclusions: Processed nerve allografts represent a viable alternative to autografts in selected cases—especially short to mid-length motor nerve defects. However, their role in facial nerve reconstruction remains insufficiently studied. Further trials are needed to address specific anatomical and functional challenges in this subgroup and to clarify long-gap efficacy. Full article

Directed grey-box fuzzers focus on testing specific target code. They have been utilized in various security applications, such as reproducing known crashes and identifying vulnerabilities resulting from incomplete patches. Distance-guided directed fuzzers calculate the distance to the target node for each node in a CFG or CG, which has always been the mainstream in this field. However, the distance can only reflect the relationship between the current node and the target node, and it does not consider the impact of the reaching sequence before the target node. To mitigate this problem, we analyzed the properties of the instrumented function’s call graph after selective instrumentation, and the complexity of reaching the target function sequence was estimated. Assisted by the sequence complexity, we proposed a two-stage function call sequence-based seed-scheduling strategy. The first stage is to select seeds with a higher probability of generating test cases that reach the target function. The second stage is to select seeds that can generate test cases that meet the conditions for triggering the vulnerability as much as possible. We implemented our approach in SEZZ based on SelectFuzz and compare it with related works. We found that SEZZ outperformed AFLGo, Beacon, WindRanger, and SelectFuzz by achieving an average improvement of 13.7×, 1.50×, 9.78×, and 2.04× faster on vulnerability exposure, respectively. Moreover, SEZZ triggered three more vulnerabilities than the other compared tools. Full article

Deposition of intramuscular fat (IM), also known as marbling, is the deciding factor of beef quality grade in the U.S. Defining molecular mechanisms underlying the differential deposition of adipose tissue in distinct anatomical areas in beef cattle is key to the development of strategies for marbling enhancement while limiting the accumulation of excessive subcutaneous adipose tissue (SAT). The objective of this exploratory study was to define the IM and SAT transcriptional heterogeneity at the whole tissue and single-nuclei levels in beef steers. Longissimus dorsi muscle samples (9–11th rib) were collected from two finished beef steers at harvest to dissect matched IM and adjacent SAT (backfat). Total RNA from IM and SAT was isolated and sequenced in an Illumina NovaSeq 6000. Nuclei from the same samples were isolated by dounce homogenization, libraries generated with 10× Genomics, and sequenced in an Illumina NovaSeq 6000, followed by analysis via Cell Ranger pipeline and Seurat in RStudio (v4.3.2) By the expression of signature marker genes, single-nuclei RNA sequencing (snRNAseq) analysis identified mature adipocytes (AD; ADIPOQ, LEP), adipose stromal and progenitor cells (ASPC; PDGFRA), endothelial cells (EC; VWF, PECAM1), smooth muscle cells (SMC; NOTCH3, MYL9) and immune cells (IMC; CD163, MRC1). We detected six cell clusters in SAT and nine in IM. Across IM and SAT, AD was the most abundant cell type, followed by ASPC, SMC, and IMC. In SAT, AD made up 50% of the cellular population, followed by ASPC (31%), EC (14%), IMC (1%), and SMC (4%). In IM depot, AD made up 23% of the cellular population, followed by ASPC at 19% of the population, EC at 28%, IMC at 7% and SMC at 12%. The abundance of ASPC and AD was lower in IM vs. SAT, while IMC was increased, suggesting a potential involvement of immune cells on IM deposition. Accordingly, both bulk RNAseq and snRNAseq analyses identified activated pathways of inflammation and metabolic function in IM. These results demonstrate distinct transcriptional cellular heterogeneity between SAT and IM depots in beef steers, which may underly the mechanisms by which fat deposits in each depot. The identification of depot-specific cell populations in IM and SAT via snRNAseq analysis has the potential to reveal target genes for the modulation of fat deposition in beef cattle. Full article

Field-based landslide mapping is a crucial task for geo-hydrological risk assessment but is often limited by the lack of integrated tools to capture accurate spatial and temporal data. This research investigates a Direct Numerical Cartography (DNC) system’s ability to capture both spatial and temporal landslide features during fieldwork. DNC enables fully digital surveys, minimizing errors and delivering real-time, spatially accurate data to experts on site. We tested an integrated approach combining the Ground Operative System for GIS Input Remote-data Acquisition (GOGIRA) with the Expert-based Multitemporal AI Detector (ExMAD). GOGIRA is a low-cost system for efficient georeferenced data collection, while ExMAD uses AI and multitemporal Sentinel-2 imagery to detect landslide triggering times. Upgrades to GOGIRA’s hardware and algorithms were carried out to improve its mapping accuracy. Field tests in Western Greece compared data to 64 expert-confirmed landslides, with the Range-R device showing a mean spatial error of 50 m, outperforming the tripod-based UGO device at 82 m. Operational factors like line-of-sight obstructions and terrain complexity affected accuracy. ExMAD applied a pre-trained U-Net convolutional neural network for automated temporal trend detection of landslide events. The combined DNC and AI-assisted remote sensing approach enhances landslide inventory precision and consistency while maintaining expert oversight, offering a scalable solution for landslide monitoring. Full article

Autonomous drone navigation presents challenges for users unfamiliar with manual flight controls, increasing the risk of collisions. This research addresses this issue by developing a multifunctional drone control system that integrates hand gesture recognition, obstacle avoidance, and 3D mapping to improve accessibility and safety. The system utilizes Google’s MediaPipe Hands software library, which employs machine learning to track 21 key landmarks of the user’s hand, enabling gesture-based control of the drone. Each recognized gesture is mapped to a flight command, eliminating the need for a traditional controller. The obstacle avoidance system, utilizing the Flow Deck V2 and Multi-Ranger Deck, detects objects within a safety threshold and autonomously moves the drone by a predefined avoidance distance away to prevent collisions. A mapping system continuously logs the drone’s flight path and detects obstacles, enabling 3D visualization of drone’s trajectory after the drone landing. Also, an AI-Deck streams live video, enabling navigation beyond the user’s direct line of sight. Experimental validation with the Crazyflie drone demonstrates seamless integration of these systems, providing a beginner-friendly experience where users can fly drones safely without prior expertise. This research enhances human–drone interaction, making drone technology more accessible for education, training, and intuitive navigation. Full article

Nature-inspired metaheuristics have proven effective for addressing complex structural optimization challenges where traditional deterministic or gradient-based methods often fall short. This study investigates the feasibility and benefits of embedding three prominent metaheuristic algorithms, the Genetic Algorithm (GA), the Firefly Algorithm (FA), and the Group Search Optimizer (GSO) embedded into the ANSYS Parametric Design Language (APDL). The performance of each optimizer was assessed in three case studies. The first two are spatial truss structures, one comprising 22 bars and the other 25 bars, commonly used in structural optimization research. The third is a planar 15-bar truss in which member sizing and internal topology were simultaneously refined using a Discrete Topology (DT) variable method. For both the FA and the GSO, enhanced ranger-movement strategies were implemented to improve exploration–exploitation balance. Comparative analyses were conducted to assess convergence behavior, solution quality, and computational efficiency across the different metaheuristics. The results underscore the practical advantages of a fully integrated APDL approach, highlighting improvements in execution speed, workflow automation, and overall robustness. This work not only provides a comprehensive performance comparison of GA, FA, and GSO in structural optimization tasks, but it can also be considered a novelty in employing native APDL routines for metaheuristic-based finite element analysis. Full article

A sufficient number of sensors installed in all structural components is a prerequisite for obtaining the full-field temporal–spatial displacement and is essential for large-scale structure health monitoring. In this paper, a novel lightweight vision-based temporal–spatial deflection measurement method is proposed to measure the full-field temporal–spatial displacement of slender structures. First, the geometric and mechanical properties of slender members are introduced as the priori information to vision-based measurement. Then, a slice template matching model is proposed by deploying a one-dimensional template matching model in every pixel column of each image frame, based on traditional digital image correlation (DIC) method. An indoor experiment was carried out to verify the proposed method, and experiment results show that measurement precision of STMM agrees well with the theory and the laser ranger, with a maximum measurement error of 0.03 pixels and the root-mean-square error (RMSE) of 0.052 mm, for transient beam deflection curve; with the correlation coefficient and coefficient of determination of 0.9994 and 0.9986, for dynamic deflection–time history curves at the middle-span point. Finally, further investigation reveals that brightness inconstancy is the source of STMM measurement error. Full article

Background: Veno-venous extracorporeal membrane oxygenation (V-V ECMO) represents a progressively adopted life-sustaining intervention worldwide, particularly in the management of acute respiratory distress syndrome. Nevertheless, data concerning the prognostic significance of fungal isolation in this setting remain unclear. This study aims (i) to assess the incidence of fungal infection and colonization in a homogeneous cohort of V-V ECMO patients, and (ii) to evaluate the association between fungal infection or colonization and 1-year mortality, with a focus on the impact of specific fungal species. Methods: All consecutive adults admitted to the Intensive Care Units of five Italian university-affiliated hospitals and requiring V-V ECMO were screened. Exclusion criteria were age < 18 years, pregnancy, veno-arterial or mixed ECMO-configuration, incomplete records and survival < 24 h after V-V ECMO placement. A standard protocol of microbiological surveillance was applied and the distinction between different fungal species were made through in vivo and vitro tests. Cox-proportional hazards models, Kaplan–Meier curves and linear logistic regressions were applied for investigating mortality. Results: Two-hundred and seventy-nine V-V ECMO patients (72% male) were enrolled. The overall fungal isolation was 41% (n. 114): 23% infections and 18% colonizations. The overall 1-year mortality, among fungal isolations, was 40%, with no different risk in case of fungal infection (26 out of 63, 41%) (aHR 0.85, 95% CI [0.53–1.37], p-value 0.505) and colonization (20 out of 51, 39%) (aHR 0.86, 95%CI [0.51–1.43], p-value 0.556), as compared to patients never detecting fungi (68 out of 165, 41%, reference). According to the isolated mycotic species, as compared to Candida sp. group (reference), the risk of death was greater when different fungal species (e.g., Aspergillus sp. and Candida sp.) were concomitantly isolated in the same patient (OR 1.17, 95%CI [1.12–11.07], p-value 0.031. Conclusions: In the overall population, 23% V-V ECMO patients recorded ‘late’ fungal infections and 18% fungal colonizations, with a similar risk of death as compared to patients never experiencing fungi during the V-V ECMO course. The detection of concomitant different fungal species was an independent risk factor for 1-year mortality. Full article

Amid urbanization, many forest farmers have migrated for work, leading to a shortage of young labor in forestry. Socialized prevention and control (SPC) measures have emerged as a new forestry model. By integrating forestland property rights theory, SPC economic principles, and collaborative disaster governance, this study compares the econometrics methods of seemingly unrelated regression (SUR) and structural equation models (SEMs) to address correlation and endogeneity issues. It aims to understand forest farmers’ willingness to pay for SPC services, purchase forest insurance, and join as forest rangers. The findings offer theoretical and practical insights that address current challenges and rationalize SPC promotion costs, filling gaps in the existing literature. The results indicate that high-quality foresters with more home-planted forests are more inclined to hire SPC companies, while better-educated farmers are less likely to purchase forest insurance. Western forest farmers, highly reliant on forests, show greater willingness to become rangers under village committee organization. Farmers organized by committees or with prevention experience suggest SPC costs around USD $65/ha and forest premiums at USD $5/ha, with high-quality farmers proposing a ranger salary of USD $190/month. Recommendations include collecting SPC funds from farmers and supplementing through local finances; enhancing the forest insurance system; monitoring SPC companies; and recruiting young, skilled rangers. Full article

This study evaluated heavy metal levels (Pb, Cd, Hg, As, Cu) and the presence of Salmonella spp. and Listeria monocytogenes in mussels and commonly consumed fishery products from the Marmara region of Türkiye. Health risks were evaluated using total hazard quotient (THQ) and hazard index (HI) values, while microbial risks in fresh and ready-to-eat (RTE) products were estimated via the Risk Ranger tool. Among 625 samples, Hg (36.96%; CI95 = 33.27–48.81), Pb (9.76%; CI95 = 7.67–12.34), and Cd (19.36%; CI95 = 16.45–22.64) exceeded permissible limits, except in crabs, which remained compliant. Anchovy, sardines, bluefish, shrimps, and octopus met EU Cd limits. Shrimps exhibited higher As_in levels than crabs (p < 0.05), while squids had significantly higher As_in than octopus but lower Pb (p < 0.05). Microbiological analysis detected Salmonella spp. in 4.00% of samples (CI95 = 2.50–6.30) and L. monocytogenes in 4.24% (CI95 = 2.70–6.59). Surmullet, bluefish, red mullet, crabs, mussels, and octopus tested negative for both, while anchovy was negative for Salmonella spp. only. THQ and HI assessments emphasized the need for environmental monitoring to mitigate heavy metal contamination. The detection of pathogens highlights the importance of stringent surveillance measures to ensure the safety of fishery products and bivalves. Full article