Search Results (187)

Background/Objectives: The aim of the present investigation was to validate the diagnostic performance of the AI-based dental cloud software Diagnocat^® AIS (Version 1.0 (UDI: 860010268018), DGNCT LLC, Miami, FL, USA) regarding the detection possibilities of seven different endodontic findings on cone-beam computed tomography (CBCT) against a multi-rater consensus reference standard, and to characterize its calibration, threshold-optimized performance and clinical utility. Methods: 358 root-canal-treated teeth from 167 CBCT scans (167 patients) were retrospectively evaluated at a single private dental practice. From initially included 383 root-canal-treated teeth from 177 patients, 358 (93.5%) were recognized by the AI tool and entered the primary analysis. Two experienced dentists with a clinical focus on endodontics independently graded each tooth and disagreements were adjudicated by a senior expert. Seven different endodontic findings were evaluated: (i) apical (periapical) lesion; (ii) short root-canal filling (apical filling end >2 mm short of the radiographic apex); (iii) voids/lacunae in the root-canal filling; (iv) missed (un-instrumented/un-filled) canal; (v) overfilled root-canal filling (apical extrusion); (vi) apicoectomy (resected root apex with or without retrograde filling); and (vii) coronal restoration with a full-coverage crown. Diagnocat^® output was binarized at the manufacturer-fixed 0.50 probability threshold; sensitivity, specificity, predictive values, accuracy, area under the curve AUC (ROC), Cohen κ and Gwet AC1 were computed with 95% cluster-bootstrap confidence intervals (cluster = scan). Threshold optimization, probability calibration, GEE-based subgroup analyses, and decision-curve analysis were pre-specified. Results: Diagnostic performance varied by finding. AUCs were 0.984 for missed canal, 0.917 for overfilled root canal, 0.902 for short root filling, 0.893 for crown, 0.864 for apical lesion, 0.857 for apicoectomy and 0.761 for voids in the root filling. Apical-lesion sensitivity rose from 33.6% for sub-millimeter lesions to ≥80% for lesion measuring 1–5 mm. Re-tuning the decision threshold raised missed-canal sensitivity from 69.6% to 97.5%. Decision-curve analysis confirmed positive benefits for missed canal and root-filling-quality findings. Conclusions: The AI tool Diagnocat^® can be recommended as a focused screening adjunct in CBCT-based endodontic interpretation for missed canals, crowns, and gross root-filling-quality flaws. Sub-millimeter apical lesions and several less common findings (resorption, instrument fragment, retrograde filling) remain outside the reliable performance envelope of the current platform. Full article

Objectives: To externally validate the Barcelona Predictive Models (BCN-PM 1 and 2) for detecting csPCa in an Ibero-American population. BCN-PM 1 was designed to reduce magnetic resonance imaging (MRI) use, whereas BCN-PM 2 aims to decrease unnecessary prostate biopsies. Methods: This prospective, multicenter study included 661 men with suspected PCa recruited in 2025 across three Ibero-American centers. All participants underwent MRI followed by targeted biopsies of lesions with the Prostate Imaging–Reporting and Data System (PI-RADS) ≥ 3, along with systematic biopsy. When PI-RADS lesions were <3, only systematic biopsies were performed. CsPCa was defined as International Society of Urological Pathology Grade Group ≥ 2. BCN-PM 1 incorporates age (years), family history of PCa (no vs. yes), prior negative prostate biopsy (no vs. yes), digital rectal examination (DRE: normal vs. suspicious), and prostate volume-derived estimation by DRE (small, median, or large). BCN-PM 2 includes age, family history of PCa, prior negative prostate biopsy, prostate volume measured by MRI (mL), and PI-RADS score (1–5). Results: The rate of csPCa detection was 53.7%. Both models demonstrated good calibration with strong agreement between predicted probabilities and observed csPCa rates. BCN-PM 1 closely followed the reference line, with minor deviations at higher predicted probabilities, whereas BCN-PM 2 showed modest departures at the extremes of risk. The area under the curve was 0.740 (95% CI 0.702–0.777) for BCN-PM 1 and 0.803 (95% CI 0.769–0.836) for BCN-PM 2 (p < 0.001). Decision curve analysis demonstrated a net benefit for both models compared with strategies of biopsy in all or no men. BCN-PM 2 showed greater net benefit than BCN-PM 1. At 95% sensitivity, BCN-PM 1 reduced MRI requests by 10.6%, while BCN-PM 2 avoided 19.4% of unnecessary biopsies. The sequential use of BCN-PM 1 and 2 resulted in a 10.6% reduction in MRI exams and a 23.1% reduction in biopsies, at the cost of missing 8.4% of csPCa cases. The performance of the biopsy improved from 53.7% to 64.0% (p < 0.001). Conclusions: BCN-PM1 and BCN-PM 2 were successfully validated in an Ibero-American population. Full article

Multi-resolution image fusion has been studied for years to solve the trade-off between temporal and spatial resolution in remote sensing instruments and has been widely applied to detect and monitor natural phenomena like floods. Despite the considerable research on this topic, the mitigation of the influence of outliers, such as cloud and shadow miscorrections, on satellite image fusion has not been fully developed. Moreover, strategies that integrate robustness, recursive operation and learned models are missing. In this paper, we design a robust recursive image fusion framework leveraging a location-aware neural network (NN) to model image dynamics. Outliers are modeled by representing the probability of contamination of a given pixel and band. An NN model trained on a small dataset provides accurate predictions of stochastic image time evolution, which improves both the accuracy and robustness of the method. A recursive solution is proposed to estimate high-resolution images by using a Bayesian variational inference framework. Experiments fusing images from the Landsat 8 and MODIS instruments show that the proposed method generally reduces the root mean square error (RMSE) and misclassification percentage of the estimated image by over $10\%$ without cloud cover and over $20\%$ with cloud cover compared with the benchmark KF algorithm. This indicates that the proposed approach is significantly more robust against cloud cover, without losing performance when no clouds are present. Full article

Automatic Music Transcription (AMT) for piano is difficult for audio-only systems due to dense polyphony, resonance, and reverberation, which lead to false positives and unstable onset decisions. We present a multimodal AMT framework that fuses Omnizart audio probability maps with visual cues from hand-skeleton tracking. A graph-based model called HandSkeletonNet estimates per-key onset probabilities from hand trajectories, and the two modalities are merged via a weighting-and-masking scheme or a compact CNN-based merger. Experiments show consistent improvements over the audio-only baseline on our self-compiled dataset, while evaluations with external datasets primarily improve frame-level sensitivity. The frame-level F1 score improved from 75.12% to 75.76% for the PianoYT dataset and from 54.68% to 57.57% for the PianoVAM dataset compared with the audio-only baseline. Our experiments also reveal limited onset-level gains under domain shift. Remaining errors are largely explained by timing/misalignment and note fragmentation in MIDI decoding, suggesting that robustness to missing hand detections and explicit temporal alignment are key directions. Full article

Optical Wireless Sensor Networks (OWSNs) have emerged as a promising solution for energy-efficient and secure data collection in free-space optical (FSO) environments. A key challenge in such networks is minimizing the decision error rate (DER) during decision aggregation at the central entity (CE). Building on earlier Time-Difference-of-Arrival (TDoA) reporting methods, this paper introduces an Angle-of-Arrival (AoA) framework for decision gathering. In the proposed scheme, sensor nodes equipped with Corner Cube Retro-reflectors (CCRs) passively communicate their local decisions, while the CE identifies such decisions based on AoA estimation. A closed-form expression for the DER is derived, incorporating false-alarm and missed-detection probabilities, and is validated through Monte Carlo simulations. Comparative evaluation against TDoA, Single Wavelength Parallel (SWP), and Multiple Wavelength Series (MWS) schemes shows that the AoA-based approach achieves consistently lower DERs, particularly in high-SNR regimes and larger node counts, closely approaching the theoretical lower bound. These results highlight AoA as a practical and scalable alternative to conventional decision-gathering methods in OWSNs. Full article

Background: Operational performance in security-force settings depends on maintaining accurate motor–cognitive and shooting performance under acute physical strain. This descriptive pilot study examined pre–post performance patterns during a fixed-order acute-load protocol and explored whether trial-level and spatial analyses identified changes beyond aggregate scores. Methods: Nineteen applicants (10 men, 9 women; 21.6 ± 1.0 years) completed two testing sequences separated by one week. All participants completed Sequence 1 first and Sequence 2 second; therefore, sequence-related observations were interpreted descriptively rather than causally. In both sequences, participants performed Hawk Eye testing, IPSC-based shooting, and the Jaciak Motor Coordination Test, with the order of Hawk Eye and shooting reversed between sequences. Primary outcomes were first-shot hit rate and Hawk Eye error count. Secondary and exploratory outcomes included shooting miss rate and time, Hawk Eye stimulus time, minimum and maximum response times, trial-level timing, spatial distributions, and cross-task coupling. Results: Heart rate increased markedly after the Jaciak test in both sequences, with end-of-test values corresponding to approximately 86–88% of age-predicted HRmax. Model-based analysis indicated lower post-load odds of a first-shot hit compared with pre-load performance. In contrast, no detectable pre–post change was observed for Hawk Eye error probability. Descriptively, first-shot hit rate decreased in Sequence 1 (62.1 ± 19.9% vs. 42.1 ± 28.2%; p = 0.029), while the decrease in Sequence 2 was smaller and not statistically significant (61.1 ± 24.5% vs. 52.6 ± 28.4%; p = 0.267). Hawk Eye error count showed no statistically detectable pre–post change in either sequence, although maximum response time decreased in Sequence 1 (p = 0.008). Trial-level and spatial analyses indicated additional temporal and location-specific patterns, but exploratory cross-task spatial associations were inconsistent. Conclusions: In this fixed-order descriptive pilot study, post-load testing was associated with lower first-shot shooting performance in this sample, whereas no statistically detectable deterioration was observed for Hawk Eye error probability. However, because the design lacked a no-load control condition and all participants completed the same sequence order, the observed pre-to-post differences cannot be attributed specifically to acute physical load. They should be interpreted as descriptive performance patterns within the implemented protocol. Full article

To improve the safety and reliability of tidal stream turbines (TSTs) under harsh marine environments, a novel probabilistic approach is proposed for blades fault detection in TSTs subject to stochastic disturbances of unknown probability distribution. On the basis of analytically analyzing the influence of blade imbalance fault on stator current signals, stationary wavelet transform (SWT) is first performed to extract multiscale time–frequency characteristics of blade faults from stator current data corrupted by non-stationary stochastic disturbances. Then an enhanced feature space is established by further computing the energy, standard deviation and kurtosis of SWT decomposition coefficients. By introducing the mean-covariance-based ambiguity set to characterize the probability distribution of feature vector in both fault-free and faulty cases, an optimal separating hyperplane for fault detection is learned using a distributionally robust optimization technique. It can achieve an optimal trade-off between the false alarm rate and the missed detection rate in a probabilistic setting, without requiring any specific distribution assumption. In this way, the proposed fault detection system is robust not only against disturbances but also against distributional uncertainties of disturbances. Finally, an experimental study based on a 0.23 kW tidal stream turbine platform is carried out to validate the effectiveness of the proposed method. Full article

Aiming to improve radar multi-target tracking (MTT) accuracy and association performance in complex scenarios involving dense clutter, missed detections, and maneuvering targets, an improved tracklet generation approach based on the expectation–maximization (EM) framework is proposed in which data association variables and motion model variables are jointly modeled as latent variables. These variables are estimated through iterative updates based on the loopy belief propagation (LBP) algorithm and the interacting multiple model (IMM) filtering and smoothing algorithms to generate high-confidence tracklets. Then, a delayed decision-making strategy based on the multi-hypothesis approach is employed to associate these tracklets into complete target trajectories. The resulting algorithm is named IMM-TrackletMHT. The performance of the IMM-TrackletMHT algorithm is evaluated and compared with several baseline algorithms in simulated scenarios under different clutter rates and detection probabilities. The simulation results demonstrate that the proposed algorithm consistently outperforms the baseline methods in terms of tracking accuracy, exhibits strong robustness to variations in the operating environment, and achieves higher computational efficiency in multi-scan measurement processing, thereby demonstrating the effectiveness and superiority of the proposed tracklet generation approach for maneuvering MTT. Full article

A neural network simulation–regression method was developed to assess the impact of sanctions on the level of administrative offences under fragmented, noisy, and short administrative time series. The study addresses the problem of quantifying and predicting changes at the offence level as a sanction size function, using detection probability, prior violation level, compliance costs, and auxiliary contextual factors. The proposed framework combines a hybrid MLP–LSTM neural network, double machine learning-based orthogonal causal estimation, the simulation-based generation of counterfactual scenarios through domain randomization, multiple imputation for missing data, debiasing procedures, and ensemble uncertainty estimation. The contribution to administrative law consists of a quantitative tool creation for substantiating and optimising sanction policy, assessing heterogeneous effects, and supporting evidence-based rulemaking and law enforcement decisions. In comparative experiments, the developed method achieved an RMSE of 8…12%, a prediction accuracy of 93…96%, an overall accuracy of 95%, a precision of 94%, a recall of 93%, and an F1-score of 93.5%, thereby outperforming contemporary econometric, simulation, causal machine learning, and predictive machine learning approaches used for sanction effect modelling. Additional verification through nonparametric statistical testing confirmed that the proposed model’s superiority over the compared algorithms is statistically significant across the main quality metrics, which strengthens the evidence for its robustness and practical value in sanction policy analysis under fragmented administrative data conditions. Full article

In challenging environments, there often exist problems of false alarms and missed detections in real-time kinematic (RTK) ambiguity resolution, which significantly reduce the reliability and availability of position information. To address these issues, a machine-learning method is proposed to conduct a correctness check on RTK ambiguity fixing, aiming to reduce the occurrences of false alarms and missed detections. The inter-system differential RTK model is adopted. Compared with the traditional RTK model, this model can provide an effective feature, namely the differential inter-system biases (DISB), to improve the accuracy of machine-learning classification. This is because when the RTK ambiguity is correctly fixed, the DISB usually appears as a stable constant. In addition to DISB, features that are strongly related to ambiguity fixing, such as the ratio value, DOP value, and residuals, are also comprehensively utilized. This method is verified by an open-source, large-scale, and diverse GNSS/SINS dataset—SmartPNT-POS. The experimental results show that, compared with the traditional method of relying solely on the empirical ratio value for ambiguity fixing verification, the missed detection probability of this method is reduced by 2%, the false-alarm probability is decreased by 29%, and the positioning accuracy is improved by approximately 7%. Moreover, compared with other features, the DISB feature provides the highest contribution rate in the machine-learning classification model. Full article

Metallic contaminants in powdered foods represent a serious safety concern. Therefore, effective detection is crucial for food safety. This study aimed to develop an electric-field-based detection system and quantitatively evaluate its performance. An alternating (+/−) electrode array (gap 1–2 mm) was designed, and resonance analysis identified 15 kHz with a 2 mm gap as the optimal operating condition. Using an IGBT-based high-voltage source, 1.35 kV was selected to ensure stable operation without partial discharge. A real-time algorithm based on a minimum current-change threshold was implemented, and detection responses to stainless steel (SUS), aluminum (Al), and copper (Cu) particles in three size classes (<0.5, 0.5–1.0, and 1.0–2.0 mm) were evaluated using hit/miss modeling and logistic regression to obtain probability-of-detection (POD) curves and limits of detection (LOD). The system achieved POD ≥ 0.9 for 1.0–2.0 mm particles; in the 0.5–1.0 mm range, observed POD values were 84%, 90%, and 68% for SUS, Al, and Cu, respectively. Safety was assessed by COMSOL-based localized heating simulation validated by infrared thermography and by ozone monitoring for real-time operation. Compared with conventional inspection approaches, the proposed system provides a compact, cost-effective architecture while reporting inspection-oriented reliability metrics (POD/LOD) for process-line deployment. Full article

Continuous-variable quantum communication (CVQC) relies on finite-window estimation of phase space moments, making receiver decisions sensitive to finite measurement resolution, calibration uncertainty, and confidence-calibrated tolerances. This paper develops a receiver-centric threat modeling framework for structured (including adversarial) physical-layer disturbances under finite-sample inference. We introduce an operational taxonomy, reconnaissance, exploratory, and denial-of-service, defined by statistical visibility relative to acceptance regions rather than by assumed physical mechanisms. Using an effective estimator space Gaussian model ${\widehat{\mathbf{r}}}^{\prime }=G\widehat{\mathbf{r}}+\mathbf{\xi }$ with additive covariance N, we show how distinct mechanisms can be observationally equivalent within finite tolerances and we propose a protocol-agnostic scalar severity coordinate $\Delta E$ based on the covariance trace. We derive ${\chi }^2$-based missed-detection expressions and a soft detectability boundary scaling as $1/\sqrt{n}$, and we corroborate the predicted $P_{\mathrm{miss}}\left(\nu \right)$ behavior via Monte Carlo simulations across representative block sizes. The resulting framework clarifies the delimitation from conventional CV-QKD excess noise parameterization and provides a structured basis for monitoring-layer design and comparative threat-taxonomy mapping. Full article

Non-intrusive load monitoring (NILM) can provide actionable feedback for demand response (DR) when direct measurements of device states are unavailable. We propose a reproducible, engineering-oriented pipeline for detecting ON/OFF states of end-use load groups from an aggregated active power time series. The method uses robust hysteresis-based labeling with adaptive thresholds derived from the median and median absolute deviation, followed by compact feature engineering restricted to global active power (GAP). After removing collinear features (|r| > 0.98), permutation importance is used to retain informative predictors. Probabilistic binary classifiers (LGBM, Histogram-based Gradient Boosting, XGBoost, and CatBoost) are trained for each target load, and the decision threshold is optimized via F_β to balance missed events and false alarms. A post-processing stage stabilizes predictions by smoothing probabilities and suppressing spurious triggers. Model quality is assessed with both sample-wise metrics and event-based metrics that credit the correct detection of switching intervals within a time tolerance. Experiments on the open “Individual Household Electric Power Consumption” dataset (1-min resolution, 2007–2010) demonstrate that lightweight gradient boosting models, particularly LGBM, deliver reliable and interpretable state estimates suitable for practical DR integration and edge deployment. Full article

To address the problem that the detection performance of existing spectrum sensing algorithms degrades or even fails under impulsive noise, this paper proposes a generalized energy detection-based spectrum sensing algorithm. Theoretical analysis verifies that the proposed algorithm can effectively mitigate the adverse effects of impulsive noise, realize high-precision signal detection, and enhance system reliability with fewer samples. Furthermore, through statistical theoretical analysis, the probability density function of the detection statistic is provided for both scenarios where the primary user signal is absent and present. The probabilities of false alarm and missed detection are also derived, and the threshold corresponding to a prescribed false alarm probability is determined. Finally, simulation results demonstrate the effectiveness of the generalized energy detection algorithm. Full article

Early wildfire smoke detection is critical for preventing small ignitions from escalating into large-scale fires, yet early-stage smoke plumes are often faint, low-contrast, and spatially small. When full-resolution frames are resized to satisfy fixed-input detector architectures and enable efficient batched GPU inference, these subtle cues are further diminished, leading to missed detections and unreliable scores near deployment thresholds. Existing remedies such as multi-scale inference, slicing/tiling, or super-resolution could improve sensitivity, but typically incur substantial overhead from multiple forward passes or added network components, limiting real-time use on resource-constrained platforms. To mitigate these challenges, we propose a composite multi-resolution detection framework that improves sensitivity to small smoke regions while maintaining single-pass inference. Motivated by the fact that most operational wildfire monitoring systems rely on Unmanned Aerial Vehicle (UAV) platforms and mountain-top Closed-Circuit Television (CCTV) systems surveillance, their wide-field imagery typically contains a large sky region above the horizon where early smoke is most likely to first become visible. Accordingly, crop placement is guided by a skyline prior that prioritizes this high-probability sky band while retaining the remaining scene for global context. A dynamic compositing stage stacks a global view with a high-resolution, sky-aligned band into a standard square detector input, preserving context with minimal added cost. Detections from the two views are reconciled via coordinate restoration and non-maximum suppression. For deployment, a lightweight second-stage classifier selectively re-evaluates low-confidence detections to stabilize decisions near a fixed operating threshold without retraining the detector. Compared to the baseline detector, our approach improves detection performance on the Early Smoke dataset, achieving gains of +4.6 percentage points in AP @0.5:0.95, +3.4 percentage points in AP @0.5, +2.9 percentage points in precision, +5.3 percentage points in recall, and +4.3 percentage points in F1-score. Full article