Search Results (886)

Under climate change, extreme wind events are predicted to become both more common and more severe, increasing the vulnerability of plantation forests. In February 2023, ex-tropical Cyclone Gabrielle caused widespread wind damage to radiata pine (Pinus radiata D. Don) forests across the North Island of New Zealand, providing a rare opportunity to evaluate mechanistic wind-risk modelling under extreme storm conditions. This study assessed the performance of the ForestGALES model in predicting wind damage within the Rangipo genetic accelerator trial and examined the influence of stocking and cultivation on wind vulnerability. Using detailed pre-cyclone field measurements and high-resolution unmanned aerial vehicle light detection and ranging (ULS) data, ForestGALES was parameterised for the Rangipo trial and applied at both individual-tree and stand scales. Model predictions were compared with observed post-cyclone damage using balanced area under the receiver operating characteristic curve (AUC), accounting for strong class imbalance. Wind damage was observed in 16.7% of trees, of which 10.2% showed stem breakage and 6.5% overturning. Across both spatial scales, overturning was more accurately predicted than stem breakage. At the individual-tree scale, ForestGALES showed moderate predictive skill, with balanced AUC values of 0.650 ± 0.014 for overturning, 0.595 ± 0.011 for breakage, and 0.621 ± 0.008 for total damage. Model performance was stronger at the stand scale, where discrimination was highest for overturning (AUC 0.811 ± 0.122), followed by breakage (0.693 ± 0.116) and total damage (0.623 ± 0.083). Silvicultural treatments significantly influenced predicted critical wind speeds (CWS). High-stocking treatments exhibited consistently higher CWS values and therefore greater wind firmness than low-stocking treatments, while cultivation effects were smaller but significant. Simulated reductions in stocking further demonstrated increased wind vulnerability as stocking declined, highlighting thinning as a primary determinant of wind risk. These findings demonstrate that ForestGALES can reliably discriminate wind damage at operational stand scales under extreme cyclone conditions and highlight the importance of stand structure in improving plantation resilience under increasingly storm-prone climates. Full article

This study aimed to detect growth stress at the individual-tree level in chestnut (Castanea crenata Sieb. et Zucc.) plantations using UAV-based RGB orthomosaic and multispectral imagery and to determine an optimal NDVI threshold for stress classification. UAV surveys were conducted over a 21 ha chestnut orchard located in Gongju, Chungcheongnam-do, Republic of Korea. NDVI was calculated and analyzed at the individual-tree level using multispectral imagery. Based on field observations, 100 healthy trees and 23 stressed trees were selected for statistical analysis. The mean NDVI value was 0.900 ± 0.012 for healthy trees and 0.816 ± 0.013 for stressed trees, showing a highly significant difference (p < 0.001). ROC analysis showed excellent classification performance with an AUC of 1.00. The optimal NDVI threshold determined using Youden’s index was 0.855. Independent validation in another chestnut plantation approximately 1 km away achieved high classification accuracy using the same threshold. These results indicate that UAV-based multispectral imagery combined with NDVI analysis provides an effective approach for early detection of growth stress and precision monitoring at the individual-tree level in chestnut plantations. This study provides a practical and efficient approach for the early detection of growth stress at the individual-tree level, enabling early intervention against potential declines in tree vitality and proactive management in chestnut orchards. The proposed NDVI threshold-based method offers a simple yet robust tool that can be readily applied in precision forestry and smart agriculture to support large-scale monitoring and informed management decisions for maintaining orchard productivity, enabling cost-effective early intervention at the individual-tree level, which is difficult to achieve using conventional ground-based surveys in complex mountainous orchards. Full article

The aviation industry operates in a security-sensitive environment where customer feedback may contain not only expressions of satisfaction or dissatisfaction but also threatening or violent language with potential security implications. While conventional sentiment analysis effectively captures customer opinions, it remains insufficient for identifying security-relevant linguistic cues that could signal risks requiring proactive intervention. This study addresses this gap by introducing a security-aware ambient intelligence framework for detecting violent language in airline customer reviews. This framework supports intelligent internet-based monitoring systems and real-time threat detection. We present the first annotated dataset of airline reviews specifically labeled for violent and threatening content, derived from 3629 reviews and balanced through manual resampling to achieve equal representation across positive, neutral, negative, and violent classes. The proposed framework employs VADER-based sentiment analysis for initial polarity estimation, combined with a validated annotation process to identify violent or threat-related content, followed by comprehensive feature engineering combining TF-IDF (2000 features) with text statistics and sentiment scores. We systematically evaluate individual classifiers (Random Forest, Decision Tree, SVM, Naive Bayes) against ensemble methods (Voting, Stacking, Boosting) using accuracy, precision, recall, F1-score, and ROC AUC metrics. Results demonstrate that Stacking achieves the highest raw performance (98.57% accuracy, F1-macro 0.9856), while Naive Bayes offers an optimal balance between effectiveness and computational efficiency (81.79% accuracy, F1-macro 0.8172, training time 0.03 s). This is the first dataset and framework designed for security-aware analysis of airline reviews. The selected Naive Bayes model achieves per-class F1-scores of 0.9978 for neutral, 0.7814 for negative, 0.7482 for violent, and 0.7415 for positive reviews, with a macro-average ROC AUC of 0.7123. The framework is deployed with serialized components enabling real-time prediction, supporting both single-review analysis and batch processing for integration into airline security monitoring systems. This work establishes a foundation for security-aware natural language processing in critical infrastructure contexts, bridging the gap between conventional sentiment analysis and proactive threat detection. Full article

Integrating existing tree-level information into harvester operator decision-making can significantly enhance precision forest management, particularly with respect to biodiversity preservation and climate-smart adaptation. During harvester operations, a primary challenge lies in positioning the machine with sufficient accuracy in real time to relate a priori individual-tree-level reference information to the operator. We propose a lightweight procedure using tree-to-tree matching to continuously register a real-time tree map collected from a harvester (or another mobile laser scanning system) to a precomputed reference map derived from an airborne laser scanner (ALS). We assess the robustness of the method using simulated tree maps and validate its real-world performance in experiments using a LiDAR-equipped harvester performing a thinning operation in a boreal forest. In simulations, registration was found to be robust up to a moderate tree density of approximately 1700 ha⁻¹, even when using a reference map with a lower positional accuracy and higher error rates than in our harvester experiments. Using real-world data from the thinning operation, the registration method was demonstrated to successfully mitigate meter-scale positioning drifts remaining in the LiDAR-inertial trajectory. After the continuous registration procedure, the positioning error was reduced to the level of 0.5 m, constrained by the accuracy of the prior map derived from sparse ALS data with ∼5 transmissions/m². Importantly, the registration procedure was shown to update in real time (at most 20 ms update time for stands with densities of at most 2000 ha⁻¹, after an initial computational phase. Notable features of the registration procedure are its low memory consumption, fast runtime and capacity to accurately position the harvester despite LiDAR-inertial positioning drift. While these results demonstrate the potential for real-time operation, full implementation requires the development of real-time tree detection and estimation of tree attributes. Full article

Chronic stress is a time-dependent condition characterized by sustained dysregulation across neural, autonomic, and endocrine systems, with important consequences for both health and socioeconomic outcomes. Unlike acute stress, which is typically characterized by short-lived physiological activation, chronic stress reflects an accumulated allostatic load and a longer-term recalibration of stress response systems. Recent advances in physiological sensing and artificial intelligence (AI) have supported the development of computational approaches for chronic stress detection using electroencephalography (EEG), heart rate variability (HRV), photoplethysmography (PPG), electrodermal activity (EDA), and wearable multimodal platforms. This narrative review examines current AI-based studies through three main inferential paradigms: resting baseline dysregulation, longitudinal physiological monitoring, and reactivity-based inference. Across modalities, classical machine learning (ML) methods, particularly support vector machines (SVMs) and tree-based ensembles, remain the most commonly used approaches, largely because available datasets are small and most pipelines still depend on engineered features. Deep learning (DL) methods are beginning to emerge, but their use remains constrained by the lack of large, standardized, longitudinal datasets specifically designed for chronic stress research. Major challenges include ambiguity in stress labeling, limited longitudinal validation, circadian confounding, inter-individual variability, and small cohort sizes. Future progress will depend on standardized datasets, biologically grounded multimodal integration, hybrid baseline-reactivity modeling, adaptive personalization, and more interpretable AI systems. Greater emphasis is also needed on clinical relevance and generalizability if AI-based chronic stress monitoring is to move beyond experimental settings. Full article

Major depressive disorder (MDD) is a prevalent mental health condition that inflicts a high burden on individuals and healthcare systems. There is a clinical need to detect MDD relapse practically and effectively to improve treatment outcomes for patients. To address this, we developed a smart monitoring system using an Artificial Intelligence (AI) approach to estimate MDD severity and relapse risk from patients’ smartphone behavioral data (i.e., digital phenotyping). Thirty-five MDD patients were recruited from the Institute of Mental Health in Singapore, who installed the smartphone study app Sallie. Their symptoms were quantified using the Hamilton Depression Rating Scale (HAMD-17) at the start of the trial, and every 30 days after over 3 months. The app collected behavioral data such as activity, activity type, and GPS location used to train AI models such as logistic regression, decision trees, and random forest classifiers. We found that passive data collection continued for most participants (up to 79% retention rate) after 3 months. We also used five-fold cross-validation to predict HAMD-17 severity ranging from two to four classes and the relapse status, achieving 91%, 88%, and 78% accuracies for two to four classes, respectively, and a relapse prediction accuracy of 86% whereby four patients relapsed during the study. Additionally, anxiety factors within the HAMD-17 were significantly predicted (Pearson correlation coefficient = 0.78, p = 1.67 × 10⁻¹⁴). These results demonstrate the promise of using smartphone behavioral data to estimate depressive symptoms and identify early indicators of relapse. Full article

Diabetes mellitus is a chronic condition that frequently leads to severe complications that are difficult to detect in their early stages using conventional clinical monitoring. This paper presents a data-driven framework for predicting multiple diabetes-related complications using structured electronic health record data while ensuring clinically meaningful explainability. The proposed approach adapts a pretrained electronic health record foundation model to operate on static patient data and integrates it with classical machine learning baselines to address class imbalance, feature sparsity, and interpretability challenges. A multi-label prediction setting covering eight common diabetes complications is evaluated using a real-world dataset from a regional diabetes center in the United Arab Emirates. Synthetic data generation and clinical constraint enforcement are applied to improve robustness for underrepresented outcomes, while feature selection is guided by model importance and attribution-based explanations. The best-performing configuration, a weighted ensemble combining a low-rank adapted Hyena-based foundation model with a tree-based predictor, achieved an average F1-score of 0.77, an average recall of 0.85, and an example-based F1-score of 0.71, outperforming all individual models. In addition, this ensemble produced the most stable explanations under input perturbations, indicating improved consistency of dominant clinical risk drivers. These results demonstrate that explainable foundation model-based ensembles can deliver accurate, robust, and clinically transparent risk prediction for diabetes complications. Full article

Major Depressive Disorder (MDD) is a significant global health burden that contributes to disability and reduced quality of life. Its impact extends beyond individuals, placing emotional, social, and economic strain on families and healthcare systems worldwide. Despite its prevalence, MDD remains widely misunderstood, with limited mental health literacy and persistent stigma often preventing individuals from seeking help. This research explored the prediction of MDD utilising social media data via Natural Language Processing (NLP), Machine Learning (ML), and explainable Machine Learning (xML) techniques. The research aimed at identifying depressive indicators on X (formerly Twitter) and developing interpretable models for depression risk detection. The study’s methodology followed the Cross-Industry Standard Process for Data Mining (CRISP-DM) framework to ensure a systematic approach to data analysis. Data was collected via X’s API and processed using regex-based noise removal, normalisation, tokenisation, and lemmatisation. Symptoms were mapped to DSM-5-TR criteria at the post-level, with user-level MDD risk assessed based on symptom persistence over a two-week period. Risk levels were classified as No Risk, Monitor, and High Risk to facilitate early intervention. Six ML models were trained and tested, while the Synthetic Minority Over-sampling Technique (SMOTE) was applied to mitigate class imbalance. The dataset was partitioned into training and testing sets using an 80:20 split. ML models were evaluated, and the Extreme Gradient Boosting model outperformed the others. Extreme Gradient Boosting achieved an accuracy of 0.979, F1-score of 0.970, and ROC-AUC of 0.996, surpassing benchmark results reported in prior studies. Explainability techniques, such as LIME and tree-based feature importance, enhance model transparency and clinical interpretability. Depressed mood consistently emerged as the highest-weighted predictor across different models. The findings highlight the value of aligning ML models with validated diagnostic frameworks to improve trustworthiness and reduce false positives. Future research can expand beyond text-based analysis by incorporating multimodal features to broaden diagnostic depth. Full article

We conducted a cost-effectiveness analysis of a universal newborn screening (NBS) program for infantile-onset Pompe disease (IOPD) compared with clinical identification in newborns. The analytical model combined a decision tree and a Markov model. The incremental cost-effectiveness ratio (ICER) was estimated over a lifetime horizon, applying a 2% annual discount rate from the public healthcare payer’s perspective. In a cohort of 727,288 individuals, 2.4 patients were expected to have IOPD. The cumulative quality-adjusted life years (QALYs) gained per patient were estimated to be 7.9 when clinically diagnosed and treated with enzyme replacement therapy, and 28.9 when identified through universal NBS. The ICER was 174 million JPY per QALY. Sensitivity and scenario analyses indicated that the parameters most affecting the ICER were the NBS test cost, the quality-of-life value for ambulatory patients, the prevalence of IOPD, and the cost of enzyme replacement therapy. Although considerable uncertainty exists in the analysis, the findings suggest that implementing NBS solely for detecting infantile-onset cases poses challenges in terms of cost-effectiveness, primarily due to the rarity of the disease and the high costs associated with testing and treatment. Full article

Active non-line-of-sight (NLOS) human detection aims to infer the presence of hidden individuals by analyzing indirectly reflected photons between a relay surface and occluded targets. In this study, a single-photon avalanche diode (SPAD) and time-correlated single-photon counting (TCSPC)-based acquisition system were used to measure time–photon waveforms in controlled NLOS environments designed to represent post-disaster rubble scenarios. Although the effective temporal resolution of the system is limited by the detector timing jitter and laser pulse width, the recorded transient signals retain distinguishable intensity and temporal delay patterns associated with the primary and secondary reflections. To construct a representative dataset, measurements were collected under varying subject poses, orientations, and surrounding object configurations. The recorded signals were processed using a unified preprocessing pipeline that included normalization, histogram shaping, and signal windowing. Three machine learning models, namely, Convolutional Neural Network, Gated Recurrent Unit, and Random Forest, were trained and evaluated for human presence classification. All models achieved full sensitivity in detecting human presence; however, notable differences emerged in the classification of human-absent scenarios. Among the tested approaches, random forest achieved the highest overall accuracy and specificity, demonstrating stronger robustness to statistical variations in time–photon histograms under limited photon conditions. These results suggest that tree-based classifiers capture amplitude distribution patterns and temporal dispersion characteristics more effectively than deep neural architectures under the present acquisition constraints. Overall, the findings indicate that low-cost SPAD-based NLOS sensing systems can provide reliable human detection in indirect-observation scenarios. Full article

A monolithic digital Silicon Photomultiplier (SiPM) featuring 1024 microcells with a 30-micrometer pitch and a 50% fill factor has been designed in a 110-nanometer CMOS image sensor technology. The device under consideration integrates both SPAD sensors and front-end electronics in the same substrate. It can count up to 1024 photons in less than 22 ns, while assigning timestamps to the first and last detected photons with a time resolution of less than 100 ps. A parallel counter structure combined with a fast adder tree provides photon counting in digital form with low latency, whereas a carefully balanced fast NAND tree ensures a fixed-pattern time uncertainty not exceeding 26 ps. The architecture incorporates in-pixel memory for individual cell disabling and configurable thresholding on the timing signal for noise mitigation. In order to optimize the fill factor, a part of the electronics is placed outside the array, while the most sensitive elements of the timing and counting circuits are laid out close to the sensor, in the SPAD array. A serial readout is employed to provide a single output connection per SiPM, thereby simplifying system integration. Full article

Rising temperatures and droughts are triggering forest die-off in climate warming hotspots such as the Mediterranean Basin. UAVs equipped with LiDAR and multispectral sensors offer a powerful tool for surveys of tree vigor at landscape level. We used UAV-acquired LiDAR data and multispectral camera imagery to segment individual tree crowns, classify species, and assess the health status in two drought-affected forests in northeastern Spain: a mixed Pinus pinaster–Quercus ilex forest and a Pinus halepensis forest. Individual trees were segmented and classified using object-based image analysis with the Random Forest algorithm incorporating spectral, structural, and topographic variables. Greenness indices (NDVI and EVI) were analyzed in relation to crown height, topography (slope and elevation) and solar radiation, and their interactions. Analyses showed satisfactory crown segmentation (F-Score = 0.85–0.86) and species classification (Overall accuracy = 0.86–0.99), though distinguishing spectrally similar classes remained challenging. Taller P. pinaster trees exhibited higher NDVI, while taller P. halepensis displayed higher NDVI values in dense neighborhoods and on gentle slopes. These findings highlight the potential of high-resolution UAV-based remote sensing for effective near-real-time detection and attribution of forest die-off. Future research should aim to improve algorithm accuracy and better integrate field-based validation across different forest types. Full article

The accurate estimation of aboveground biomass (AGB) in Moso bamboo forests is critical for assessing their carbon sequestration potential and supporting sustainable management. Satellite-based approaches are often constrained by signal saturation and mixed-pixel effects, whereas Unmanned Aerial Vehicle (UAV) imagery enables precise individual tree detection, overcoming these limitations. In this study, we propose a stand density (SD)-driven AGB estimation framework using high-resolution UAV RGB imagery. Individual bamboo positions were extracted using the Revised Local Maximum (RLM) algorithm, which achieved an optimal accuracy at a 2.5 m sampling interval (OA = 82.20%). Using 85 ground-truth plots, we developed six SD-AGB models and evaluated them via 10-fold cross-validation and independent UAV validation (10 plots). The Artificial Neural Network (ANN) model outperformed others, with strong calibration (R² = 0.94, RMSE = 3.78 Mg/ha), robust cross-validation (R² = 0.84 ± 0.06, RMSE = 5.24 ± 0.67 Mg/ha), and reliable independent validation (R² = 0.87, RMSE = 4.56 Mg/ha). Spatial mapping revealed a total of 14,190 bamboo plants with an average AGB of 32.80 Mg/ha. This UAV-based SD-AGB framework provides a robust, scalable, and cost-effective tool for precise biomass estimation, supporting sustainable bamboo forest management and carbon sequestration strategies and progress towards SDG 15. Full article

Ash dieback is an often-fatal disease caused by the fungus Hymenoscyphus fraxineus (T. Kowalski) Baral, Queloz & Hosoya. It emerged in Europe during the 1990s and poses a substantial threat to ash populations. In Hungary, symptoms were first detected on common ash (Fraxinus excelsior L.) in 2008. The disease also severely impacts another native species, the narrow-leaved ash (Fraxinus angustifolia Vahl). An effective strategy for counteracting ash decline is to identify and utilize sources of tolerance. We are monitoring the health status of the selected trees that demonstrate low susceptibility (plus trees) and conducting molecular genetic studies to enable their genetic characterization and individual identification using 16 nuclear microsatellite (nSSR) markers. The PCoA (Principal Coordinates Analysis) separated the eight assessed groups into two distinct clusters based on the taxonomic traits. Based on the Structure analysis results, K = 2 was the most probable cluster number. Hybridization was also indicated in the case of several individuals across various groups. We intend to incorporate the results in the establishment of seed orchards using the selected plus trees, considering the taxonomical, geographical, and genetic distinctiveness of the different groups. Full article

Accurate detection of individual trees is essential for urban forest management and ecological assessment, yet remains challenging due to the heterogeneous backgrounds, variable sizes of tree crowns, and significant variations across urban scenarios. To address these issues, we propose Tree-SAM, a city-scale individual tree detection architecture built upon the visual foundation model Segment Anything Model (SAM) and equipped with three task-specific modules, i.e., Cross-Correlation Feature Backbone (CCFB), Hierarchical Instance Aggregation Neck (HIAN), and Context-Aware Adaptation Head (CAAH). These modules synergistically fuse general semantics with fine-grained structural cues, enable multi-scale feature aggregation, and adaptively refine predictions based on specific scene contexts. On the GZ-Tree Crown dataset, Tree-SAM achieves F1-scores of 0.762, 0.732, and 0.830, with corresponding AP@50 values of 0.478, 0.454, and 0.526 in the forest, mixed, and urban scenarios, respectively, consistently ranking first across all scenes and demonstrating strong adaptability to diverse intra-city landscapes. Additional evaluations on the BAMFORESTS dataset and the SZ-Dataset further confirm its robustness across varied geographic contexts. Tree-SAM provides a reliable, automated framework for large-scale urban tree mapping, providing reliable data support for urban forest management, carbon stock estimation, and ecological assessment. Full article