Search Results (567)

Autonomous inspection of discrete obstacles (e.g., tree trunks in orchards and forests) requires UAVs to visit every target with proper observation distance and heading, while simultaneously exploring the unknown environment. Existing space-guided exploration methods focus on eliminating unknown space and are inherently agnostic to the inspection targets themselves, leading to incomplete coverage and redundant traversal. We observe that the obstacles themselves encode the spatial topology of the environment and can serve as natural planning anchors. Based on this insight, we propose ORACLE, an Object-centric Autonomous Coverage Exploration framework that shifts the planning paradigm from space-guided to target-guided exploration. ORACLE integrates: (1) an online target detection and persistent identification module via occupied-voxel connected component labelling, (2) a density-aware global coverage planner that modulates ATSP costs to prioritize target-dense regions, and (3) a target-guided local planner that replaces frontier viewpoints with direct obstacle observation points in a Sequential Ordering Problem formulation. Experiments in two point-cloud environments reconstructed from real-world forests with contrasting tree densities (Environment I: 50 trunks, $\overline{n}=1.56$; Environment II: 70 trunks, $\overline{n}=2.19$; both with non-uniform spacing) show that ORACLE achieves $98.8\%$ and $99.7\%$ target coverage compared to $22.7\%$ and $25.1\%$ for the space-guided baseline, while reducing the mission overhead ratio from $202.9\%$ to $129.2\%$ (Environment I) and from $176.8\%$ to $126.6\%$ (Environment II). Ablation studies confirm that zone reactivation is the decisive factor for coverage completeness ($-18.8$ and $-17.2$ percentage points when disabled in Environments I and II, respectively) and that density weighting improves path efficiency. Full article

Objectives: The rise in the frequency of mental health concerns (depression and suicide) expressed on social media calls for reliable, explainable, and efficient computational methods for mental health surveillance. In this paper, we propose an interpretable framework for text-based detection of post- and community-level mental health risk on social media. Methods: The framework combines (i) Secretary Bird Optimization (SBO) for feature selection of informative linguistic and psychological features, (ii) a BERT (Bidirectional Encoder Representations from Transformers)—CNN (Convolutional Neural Network) model for post-level reasoning, and (iii) a Graph Neural Network (GraphSAGE) for community-level reasoning. The graph is estimated based on semantic similarity between posts and author relations, instead of social interactions (e.g., mentions, replies) between authors. We use SHAP and LIME for model interpretability, uncertainty, and calibration analysis to evaluate the trustworthiness of predictions. Results: The model delivers 93.1% accuracy, 0.91 F1-score, and 0.944 ROC-AUC on the eRisk and CLPsych datasets using a strict user-disjoint validation strategy. SBO lowers the number of features by about 38%, leading to better generalization. The graph-based model enables improved learning of post and user representations by capturing relational dependencies. Conclusions: Our approach offers an explainable and robust means of detecting mental health risk from text. Graph-based representations of semantic and authorship interactions enable community-level analyses, while interpretability and uncertainty estimation facilitate possible human-in-the-loop decision-making. This research does not explicitly consider a human-in-the-loop experiment. Full article

We study Membership Inference Attack (MIA) risk in Federated Learning through a two-stage lens that separates (i) whether a target client’s contribution is detectable after aggregation and system noise (Stage I: Signal Survival) from (ii) whether a surviving contribution induces a generative membership score change attributable to the target’s private data (Stage II: Signal Attribution). Stage I models aggregation as a target–background decomposition and shows that detectability hinges on target–background alignment, which can induce cancellation. Stage II connects the surviving target component to a generative MIA score via a local path representation and Lipschitz/smoothness bounds, avoiding architecture-specific assumptions. Our analysis reveals that the leading attribution term is governed by the alignment between the target update and the score geometry of the target data at an appropriate baseline. We validate the theoretical bounds and illustrate risk trajectories across several scenarios. Full article

Objectives: Real-world information on relapsed and/or refractory multiple myeloma (rrMM) clinical management in Portugal is scarce. The CharisMMa Portugal study aimed to characterize rrMM patients through socio-demographic and clinical parameters and describe treatment patterns. Methods: This was an observational, cross-sectional, multicenter study with 62 rrMM patients routinely treated at seven hospitals in Portugal. Data were collected from medical records during clinical appointments (2020–2022) after written informed consent was obtained (ClincialTrials.gov ID-NCT04135963). Patients who were diagnosed with a symptomatic MM episode in the 6 months prior to study initiation and who received treatment before their last relapse episode were enrolled. Results: Most patients were male (54.8%) and living with relatives (90.3%), and almost 50% were independent. Roughly 70% of patients were classified as Revised MM International Staging System (R-ISS) Stage II at diagnosis, with a mean age of 65.76 (±9.24) years old. Most common SliM-CRAB (SLIM: sixty percent or more clonal plasma cells in the bone marrow (S), light chain ratio ≥100 (Li), and MRI-detected focal lesions (M); CRAB: hypercalcemia (C), renal insufficiency (R), anemia (A), and bone lesions (B)) signs were bone lesions (59%), and 62.9% of the patients had at least one comorbidity. At study initiation, 70.5% of patients were on second-line treatment, with monoclonal antibodies and proteasome inhibitors (PIs) + immunomodulators (IMiDs) as leading agents. Triplet regimens were the most common across all lines, while oral and oral + subcutaneous were the most prevalent routes of administration. Conclusions: Triple treatment combinations are common in rrMM management, with PIs and IMiDs frequently used, especially in first-line settings. The use of oral formulation is substantial, suggesting a step toward more patient-centric options. This characterization underscores the complexity of rrMM management and should inform stakeholders of strategies to standardize patient care across reference centers in Portugal. Full article

Financial forecasting research often prioritizes methodological sophistication over the authenticity of underlying training data. This study quantifies the “estimation–reality divide” by comparing models trained on estimated quarterly data versus genuine, re-stated SEC-reported cash flows. Using 244 firm-quarter observations from five large-cap U.S. technology firms (Microsoft, Apple, Amazon, Alphabet, Meta; 2011–2024), this case study shows that, within this specific set of firms, models trained on estimated data exhibit a large optimistic bias. For a state-of-the-art ensemble, this bias appears as a 43% lower error rate (4.5% vs. 7.9%) compared to the same model trained on authentic data. To address this, we introduce a forecasting framework that combines (i) a Hidden Markov Model for detecting economic regimes, (ii) models tailored to each regime (XGBoost and LSTM with attention), and (iii) a dynamic ensemble that adapts to recent performance. In realistic out-of-sample tests, our framework achieves a 7.9% error rate on authentic data, significantly outperforming standard benchmarks. We also show that a meta-learning approach reduces the data needed for a new firm by about 35% while improving accuracy by 24%. In plain terms, using real SEC data leads to more honest and useful forecasts than relying on estimated data. All claims are strictly limited to the five large-cap U.S. technology firms analyzed (Microsoft, Apple, Amazon, Alphabet, Meta). No claims of generalizability to other sectors, firm sizes, or markets are made or implied. Validation on broader samples is required before extending these findings. Full article

LiDAR-based 3D object detection is critical for autonomous driving perception. Ensuring robust sensing under adverse weather is essential for safe deployment. Current physics-based simulation methods focus on atmospheric effects but offer limited ground-level modeling, leading to domain gaps between simulated and real-world snowy data. Ground-level effects are challenging to model due to diverse physical interactions: wet surface reflectivity changes, vehicle-induced spray, and multi-layer snow scattering. This paper proposes a simulation method with more comprehensive ground-effect modeling for snowfall scenarios. Our approach introduces two modules: (i) an extended spray model with precipitation-controlled parameters that jointly models spray noise and wet ground attenuation, and (ii) a multi-layer dual-mode backscattering model that captures both diffuse and specular reflections on snow-covered ground. Both modules share a unified precipitation-driven parameterization. Higher snowfall rates simultaneously control spray generation, wet surface reflectivity, and snow accumulation depth. This design ensures physical consistency and makes the approach applicable across diverse LiDAR systems without sensor-specific tuning. Experiments on the STF dataset demonstrate consistent improvements over four state-of-the-art methods under both heavy and light snowfall. Clear-weather performance is preserved. Evaluations on roadside LiDAR further confirm generalizability to infrastructure-based scenarios. Full article

The semiconductor industry has long relied on Statistical Process Control (SPC) for yield and reliability management. In early technology nodes, classic univariate tools such as Shewhart charts, cumulative sums (CUSUM), exponentially weighted moving averages (EWMA), and the Cp/Cpk exponent could effectively monitor a finite set of key variables. However, sub-5nm and emerging 3 nm technologies have fundamentally changed the statistical environment. Advanced patterning, high-aspect-ratio etching, atomic layer deposition (ALD), chemical-mechanical polishing (CMP), and novel materials have drastically narrowed the process window. At these scales, nanometer-level deviations in critical dimensions (CD), overlay, or surface roughness can significantly impact yield. Simultaneously, modern wafer fabs generate massive amounts of high-frequency sensor data and high-dimensional metrology data. Traditional SPC assumptions—such as independence, normality, low dimensionality, and stationarity—often do not hold. Semiconductor data exhibits: (i) extremely high-dimensionality and strong intervariate correlations; (ii) a hierarchical structure encompassing fab → tooling → chamber → recipe → batch → wafer → field; and (iii) metrological delays and sampling limitations leading to incomplete and asynchronous observations. To address these challenges, this paper reviews advanced statistical methods applicable to wafer fabrication. These methods include multivariate statistical process control (MSPC) approaches such as Hotelling T² statistics, PCA/PLS combining T² and Q statistics, contribution diagnostics, time-series drift and change point detection, and Bayesian hierarchical modeling for uncertainty-aware monitoring in data-limited scenarios. Furthermore, we discuss how to integrate these methods with fault detection and classification (FDC), line-to-line monitoring (R2R), advanced process control (APC), and manufacturing execution systems (MES). This paper focuses on scalable, interpretable, and maintainable implementations that transform statistical analysis from a passive monitoring tool into an active component of data-driven fab control. Full article

Fishways play a critical role in restoring river connectivity and conserving fishery resources, yet their efficiency is often limited by mismatches between hydraulic conditions and species-specific behavioral traits. To quantify the upstream migration behavior of fish under the combined influence of flow heterogeneity and schooling effects, this study examined the endangered species L. elongata in the Yangtze River Basin. Volitional swimming behavior was tested in an open-channel flume under three spatially heterogeneous flow regimes (I: Low–Moderate–High; II: High–Moderate–Low; III: Moderate–High–Low). A video monitoring system recorded the upstream movement of solitary fish and three-individual schools. Swimming trajectories, upstream migration time, preferred flow velocities, and schooling metrics—including nearest neighbor distance (NND) and mean pairwise distance (MPD)—were analyzed. Linear mixed-effects models were employed to account for repeated measures and individual variability. Results showed that schooling behavior significantly enhanced upstream migration efficiency: schooling fish arrived at the target area on average 8.93 s earlier than solitary individuals (p < 0.01), while flow condition alone had no detectable effect on arrival time. L. elongata consistently preferred low-velocity zones (0.20–0.50 m/s) and avoided high-velocity regions (0.75–1.25 m/s), with meandering upstream trajectories predominating. NND showed no significant differences across flow conditions (p > 0.05), indicating stable schooling cohesion. However, MPD increased significantly under Flow III compared to Flows I and II (p < 0.01), suggesting that higher flow heterogeneity leads to more dispersed group spacing while overall cohesion is maintained. Distinct movement strategies were observed: solitary fish predominantly utilized boundary regions as hydraulic refuges (wall-following: 63.8–80.5%), whereas schools exhibited greater spatial exploration and reduced wall-following. These findings demonstrate that schooling enhances migration efficiency while preserving a cohesive group structure and that flow heterogeneity influences within-group spatial organization. To optimize fishway performance for L. elongata, we recommend maintaining flow velocities within 0.20–0.50 m/s. This study provides scientific guidance for hydraulic regulation in fishway design and habitat restoration, emphasizing the combined effects of flow heterogeneity and schooling behavior on migration performance. Full article

Hemorrhagic shock remains a leading cause of preventable death in trauma, yet traditional vital signs may fail to reflect early blood loss before physiological compensatory mechanisms are no longer able to maintain hemodynamic stability. The Compensatory Reserve Measurement (CRM) algorithm offers early detection capability using physiological waveforms but requires testing with emerging wearable sensor technologies for operational deployment. This study tested the Epicore Epidermal Patch for Imperceptible Care (EPIC) wearable healthcare device (WHD) for CRM-based hemodynamic monitoring during progressive central hypovolemia induced by lower-body negative pressure (LBNP) to simulate hemorrhage. Twenty participants underwent progressive LBNP while photoplethysmography (PPG) signals were recorded from EPIC sensors placed at the clavicle and triceps alongside a clinical-grade finger pulse oximeter for reference. Signal quality, heart-rate accuracy, and CRM predictions were evaluated across multiple filtering approaches. The triceps placement achieved signal quality comparable to the pulse oximeter reference when Chebyshev Type II filtering was applied, as well as high heart-rate accuracy. CRM derived from the EPIC sensor placed at the triceps tracked compensatory trends during progressive hypovolemia, but prediction magnitudes were inaccurate compared to calculated CRM values. In contrast, the clavicle placement consistently performed poorly across all measurements, regardless of the signal-processing approach. These findings support the feasibility of soft, flexible wearable sensors for continuous hemorrhage monitoring at the triceps location in operational environments where traditional finger-based pulse oximetry is impractical. Full article

Background: Diabetic nephropathy (DN), also referred to as diabetic kidney disease, represents one of the most common microvascular complications of type 2 diabetes mellitus (T2DM) and remains a leading cause of end-stage renal disease worldwide. Conventional clinical markers, including albuminuria and estimated glomerular filtration rate (eGFR), are widely used for diagnosis and staging but may have limited sensitivity for detecting early renal injury and predicting disease progression. In recent years, circulating microRNAs (miRNAs) have emerged as promising non-invasive biomarkers that reflect underlying molecular mechanisms of diabetic nephropathy and may complement traditional clinical indicators. Objective: The present study aimed to evaluate serum and urinary levels of hsa-miRNA-770-5p across different stages of diabetic nephropathy and to assess its potential diagnostic value in relation to established indicators of renal function. Methods: A total of 257 participants were included and divided into four groups: healthy controls, patients with T2DM without nephropathy, patients with T2DM and DN in CKD stages I–II, and patients with DN undergoing maintenance hemodialysis (MHD). Serum and urinary levels of miRNA-770-5p were measured using quantitative real-time polymerase chain reaction (qPCR) and analyzed using the 2^−ΔΔCt method. Statistical analyses included comparisons between groups using ANOVA, correlation analyses with renal function parameters such as eGFR and proteinuria/albuminuria, and receiver operating characteristic (ROC) curve analysis to evaluate diagnostic performance. Results: Serum levels of miRNA-770-5p were significantly elevated in patients with DN and in patients undergoing maintenance hemodialysis compared with healthy controls and patients with T2DM without nephropathy. In contrast, urinary levels of miRNA-770-5p were markedly reduced in patients with DN. Serum levels in patients with T2DM without nephropathy were slightly lower than those observed in healthy controls. Significant correlations were identified between miRNA-770-5p levels and renal function parameters, including eGFR and proteinuria/albuminuria, supporting the biological relevance of this microRNA in renal injury. ROC curve analysis demonstrated good discriminatory ability for differentiating DN from T2DM without nephropathy (serum AUC = 0.82; urine AUC = 0.79). Conclusions: hsa-miRNA-770-5p demonstrates distinct and opposite patterns in serum and urine that correlate with the severity of diabetic nephropathy. The complementary changes observed in circulating and urinary levels support the potential of miRNA-770-5p as a non-invasive biomarker that may complement conventional clinical markers and provide additional insight into the molecular mechanisms involved in the development and progression of diabetic nephropathy. Full article

Mytilus galloprovincialis is a significant indicator species due to its ability to bioaccumulate environmental pollutants, such as lead (Pb), which can hinder essential reproductive molecular processes. This study aimed to examine the effect of exposure to lead (0.5, 1.5 and 5 μg/L PbCl₂) on the state of protamine-like (PL) proteins—the primary components of sperm nuclear basic proteins—and their interaction with DNA. PL proteins were analysed using acetic acid–urea PAGE and SDS-PAGE, after which their ability to bind and protect DNA from oxidative damage was also assessed. Exposure to lead resulted in SDS-PAGE-detectable alterations of the PL, particularly at levels of 1.5 µg/L and 5 µg/L of PbCl₂ and modified their capacity for DNA-binding at all doses of PbCl₂. Experiments testing the release of PLs from sperm nuclei further confirmed this, revealing a reduced release. In addition, the ability of PL proteins to protect DNA from oxidative damage was reduced at the highest exposure dose, suggesting improper condensation of sperm chromatin. Computational analyses of human protamines in the presence of lead indicated the formation of coordination complexes with Pb²⁺ in PLI-II and PL-III, potentially impairing DNA binding. Overall, our study demonstrates that exposure to lead alters the function of PL proteins and potentially destabilises the sperm chromatin of M. galloprovincialis. This provides valuable insights into the reproductive toxicity of this metal. Full article

Filtering constitutes a critical step in the post-processing of airborne Light Detection And Ranging (LiDAR) data. Over the past decade, machine learning has emerged as a prominent methodological paradigm across numerous disciplines, attracting significant research interest in its application to LiDAR filtering. From a machine learning perspective, filtering is essentially a binary classification task that aims to discriminate between ground and non-ground points. However, the limited information inherent in point clouds often leads to the generation of highly correlated features, particularly those derived from height data, which can compromise filtering accuracy. To address this issue, feature selection becomes imperative. In this study, we employed height-based mutual information as a criterion to identify and eliminate less discriminative features for filtering. The AdaBoost (Adaptive Boosting) algorithm was adopted as the classifier for point cloud filtering. For each point, nineteen features were derived from the raw LiDAR point cloud based on height and other geometric attributes within a defined neighborhood. The performance of the proposed feature selection approach was evaluated using benchmark datasets provided by the International Society for Photogrammetry and Remote Sensing (ISPRS). Experimental results demonstrate that the method is effective and reliable. After removing three selected features, the average kappa coefficient improved, along with a reduction in three categories of error, although a slight increase in Type II error (0.15%) was observed. Full article

Background and Objectives: Caries in primary teeth are characterized by rapid and often asymptomatic progression, with early dentin involvement and potential extension to the pulp. Untreated lesions may lead to complications that affect the development of the permanent dentition. The aim of this prospective study was to evaluate the clinical and biological effectiveness of chemo-mechanical controlled caries excavation using Brix 3000 compared to conventional treatment in primary molars over an 18-month follow-up period. Materials and Methods: A total of 82 children aged 4–7 years were included, each presenting with at least one carious lesion in a primary molar classified as International Caries Detection and Assessment System (ICDAS II) code 05 or 06. The carious lesions were divided into two groups according to the method of excavation: Group 1 (control), which contained 40 lesions treated with conventional bur excavation, and Group 2, which contained 42 lesions treated with chemo-mechanical excavation using Brix 3000. In all cases, excavation was controlled using a fluorescence-based device (ProFace). Clinical performance was evaluated using an assessment protocol adapted from the FDI (Fédération Dentaire Internationale) clinical criteria for the evaluation of direct and indirect restorations, with particular emphasis on biological outcomes. Follow-up examinations were performed after 1 week and 1, 3, 6, 12, and 18 months, and included radiographic evaluations. Results: After 18 months, chemo-mechanical caries excavation with Brix 3000 demonstrated a biological success rate of 100%, with no reported acute symptoms or complications. Esthetic criteria showed a success rate of 65% at 18 months, while anatomical and functional criteria demonstrated success rates of 95% and 98%, respectively. In the conventional bur excavation group, biological success reached 100%, while the esthetic, anatomical, and functional success rates were 61.3%, 93.5%, and 100%, respectively. No significant differences were observed between groups (p > 0.05). Conclusions: Chemo-mechanical controlled caries excavation using Brix 3000 represents a clinically effective and biologically reliable alternative to conventional caries excavation for the treatment of carious lesions in primary molars. Full article

Can machine learning generate statistically validated alpha in equity markets while adapting to changing market conditions? This study addresses this question by proposing a regime-aware LightGBM framework conditioned on market regimes detected via a rolling Hidden Markov Model, eliminating look-ahead bias. Backtested on 51 NASDAQ-100 constituents (2015–2026), the strategy achieved a portfolio Sharpe ratio of 1.18 (95% CI: [0.53, 1.84]) and outperformed four baseline models. The key findings include the following: (i) cross-asset features (Bitcoin as a leading indicator) contribute the most predictive value; (ii) macroeconomic indicators outweigh traditional technical indicators for high-beta stocks; (iii) the model autonomously adapts its decision logic across regimes, shifting from mean reversion in bear markets to risk appetite monitoring in bull markets. While block bootstrap tests confirm statistical significance ($p<0.001$), the Deflated Sharpe Ratio (0.69) does not reach formal significance after multiple testing correction—an honest finding we report transparently. Full article

For Autonomous Vehicles (AVs), recognizing traffic lights and signs is critical for safety because perception errors directly affect navigation decisions. Real-world disturbances such as glare, rain, dirt, and graffiti, as well as digital adversarial attacks, can lead to dangerous misclassifications. Current research lacks (i) temporal continuity (stable detection across consecutive frames to prevent flickering misclassifications), (ii) multi-field-of-view (FoV) sensing, and (iii) integrated defenses against both digital and natural degradation. This paper presents two principal contributions: (1) a three-layer defense framework integrating feature squeezing, inference-time temperature scaling (softmax $\tau$ = 3 without distillation training), and entropy-based anomaly detection with sequence-level temporal voting; (2) a 500 sequence dual-FoV benchmark (30k base frames, 150k with perturbations) from aiMotive, Waymo, Udacity, and Texas sources across four operational design domains. The unified defense stack achieves 79.8% mAP on a 100-sequence test set (6k base frames, 30k with perturbations), reducing attack success rate from 37.4% to 18.2% (51% reduction) and high-risk misclassifications by 32%. Cross-FoV validation and temporal voting enhance stability under lighting changes (+3.5% mAP) and occlusions (+2.7% mAP). Defense improvements (+9.5–9.6% mAP) remain consistent across native 3D (aiMotive, Waymo) and projected 2D (Udacity, Texas) annotations. Preliminary recapture experiments (n = 15 scenarios) show 2.5% synthetic–physical ASR gap (p = 0.18), though larger validation is needed. Code, models, and dataset reconstruction tools are publicly available. Full article