Search Results (23,694)

Early diagnosis is essential for the effective management of cardiovascular diseases (CVDs). Although conventional auscultation is the primary screening method, its reliance on subjective interpretation and susceptibility to clinical background noise have positioned phonocardiogram (PCG) analysis as a key diagnostic tool, making reliable automated interpretation a pressing necessity. Automated pipelines have evolved from handcrafted-feature machine learning to deep learning and transformer-based architectures, but the latter often depend on heavy time-frequency preprocessing and large parameter counts, inflating computational cost and increasing the risk of overfitting on limited or noisy clinical data. We propose MTENet, a Multi-representation Time-series Evidential Network that models a phase-enhanced one-dimensional PCG waveform through a bidirectional Mamba state-space encoder, capturing long-range temporal dependencies with linear-time complexity. Recordings are prepared by a label-independent, record-internal stage that combines an adaptive FFT filter bank with an automatically estimated cardiac-phase gain and returns a waveform of unchanged length and sampling rate, so that the model input remains a time-series rather than a fixed feature representation. The Mamba encoder forms one of three parallel branches, alongside an implicit neural representation (INR) branch for continuous signal modelling and a Mel-spectrogram branch computed on-the-fly within the network for spectral structure. The three streams are merged by a softmax-gated fusion. A multi-scale convolutional stem captures local transient structure, while the bidirectional Mamba encoder models longer-range cardiac rhythm. With approximately 2.58 million parameters, the model captures intricate temporal patterns while distributing representational responsibility across complementary streams. Under record-grouped four-fold validation on the primary HLS-CMDS corpus, MTENet attained an accuracy of 0.9816, a balanced accuracy of 0.9812, and an AUROC of 0.9938 under a leakage-free, record-level nested protocol in which the training epoch is selected on an inner-validation split drawn only from the training partition, so that the outer evaluation fold never informs model selection. Within-dataset evaluation on CirCor DigiScope 2022 yielded an AUROC of 0.9698. Full article

Data association for space-based radar multi-target tracking is a significant challenge in dense clutter and complex scenarios, often leading to tracking errors and target loss. This paper proposes a classification-aided message-passing algorithm for space-based radars (CA-MP-SBRs), which employs a hybrid mean-field and belief-propagation (MF-BP) framework enhanced with target classification information. We leverage high-resolution range profile (HRRP) data, which are processed by a convolutional neural network (CNN) to classify targets. The classification output is then seamlessly integrated into the MF-BP framework to jointly infer target kinematic states, visibility states, data association, and class index. Simulation results demonstrate that incorporating HRRP-based classification improves data-association reliability and tracking accuracy. Specifically, compared with Classification-aided Gaussian mixture probability hypothesis density (CA-GM-PHD), the proposed CA-MP-SBRs reduces the average root mean square error (RMSE), optimal subpattern assignment (OSPA), and generalized optimal subpattern assignment (GOSPA) by 33.4%, 31.7%, and 32.1%, respectively, while the corresponding reductions relative to classification-aided labeled multi-Bernoulli tracking for space-based radars (CA-LMB-SBRs) are 23.2%, 25.6%, and 30.3%. These results confirm the effectiveness of CA-MP-SBRs for tracking diverse target types in dense-target scenarios. Full article

Earthquakes remain among the most destructive natural hazards, necessitating reliable monitoring and early warning systems for effective risk mitigation. Recent advances in machine learning (ML) and deep learning (DL) have significantly improved seismic signal analysis, enabling more accurate event detection, phase picking, classification, and magnitude estimation. This study presents a systematic review of ML- and DL-based approaches for earthquake monitoring, with particular emphasis on Distributed Acoustic Sensing (DAS) as an emerging technology for high-resolution, real-time seismic observation. Following the PRISMA 2020 guidelines, a systematic literature search was conducted across Scopus, Web of Science, IEEE Xplore, and Google Scholar, yielding 252,223 initial records. After applying the predefined publication period, removing duplicate records, conducting relevance screening, and performing eligibility assessment, 138 peer-reviewed studies published between 2021 and 2025 were retained for detailed analysis and synthesis. The review reveals a significant transition from conventional signal-processing techniques to advanced artificial intelligence-based approaches, including Convolutional Neural Networks (CNNs), Long Short-Term Memory (LSTM) networks, Bidirectional Long Short-Term Memory (BiLSTM) networks, Transformer-based architectures, hybrid models, and Bayesian learning methods for uncertainty quantification. The findings further demonstrate that Distributed Acoustic Sensing (DAS) has emerged as a transformative sensing technology because of its dense spatial coverage, high spatial resolution, and continuous monitoring capability. However, several challenges remain, including the lack of standardized datasets, limited model generalization across diverse geological settings, insufficient model interpretability, high computational complexity, and the limited integration of uncertainty-aware approaches for real-time seismic monitoring. This review identifies these critical research gaps and highlights promising future research directions, including multimodal data fusion, interpretable artificial intelligence, physics-informed learning, self-supervised learning, and robust uncertainty quantification for next-generation intelligent seismic monitoring systems. Unlike previous review studies that primarily focus on individual machine learning techniques or conventional seismic monitoring, this review provides a comprehensive and systematic synthesis of recent advances in machine learning, deep learning, and Distributed Acoustic Sensing (DAS), identifies current research gaps, and offers practical recommendations to guide future research on intelligent earthquake monitoring systems. Full article

Rapid and reliable assessment of structural damage following disasters is critical for prioritizing rescue operations. In this study, we present a unified deep learning framework for building damage assessment from satellite imagery. The proposed approach integrates segmentation-driven feature construction, morphological processing, cost-sensitive learning, and cross-disaster evaluation to support robust performance under limited and imbalanced data conditions. The framework combines an adapted U-Net for building localization with a hybrid convolutional neural network (CNN)-deep neural network (DNN) classifier for damage-level prediction and evaluates transferability across disaster events, geographic regions, and sensing conditions. The proposed method is evaluated on selected events from the xView2 Building Damage Assessment (xBD) and BRIGHT datasets, using optical imagery from xBD and pre-disaster optical and post-disaster Synthetic Aperture Radar (SAR) imagery from BRIGHT. Despite the limited and highly imbalanced event-specific samples, the framework achieves a mean cross-validation macro-F1 score of 70% and a maximum fold-level score of 77% on the Mexico earthquake subset of xBD and up to 98% on earthquake-related events in BRIGHT. Cross-validation characterizes performance variability across source-image-grouped data partitions, while cross-disaster evaluation reveals event-dependent transferability and provides a preliminary indication that structural domain similarity may be related to transfer performance. Although the evaluation is constrained by data availability, the results indicate that lightweight, cost-sensitive deep learning frameworks may support auxiliary post-disaster screening and decision support in resource-constrained scenarios. This study highlights both the potential and the remaining challenges of deploying artificial intelligence (AI) for rapid post-disaster assessment. Full article

With the increasing application of inverter-fed permanent magnet synchronous, motors (PMSMs) in industrial and intelligent energy systems, reliable fault detection and diagnosis (FDD) has become increasingly important for ensuring operational safety and system reliability. However, conventional single-sensor-based approaches usually exhibit limited robustness under varying operating conditions due to measurement noise, load fluctuations, and incomplete fault information. Therefore, multi-sensor information fusion has attracted increasing attention in PMSM fault diagnosis because it can provide complementary information from different sensing sources. This paper proposes a hybrid deep learning framework for multi-class PMSM fault diagnosis, integrating successive variational mode decomposition (SVMD)-based signal denoising, parallel temporal and spatial feature extraction using temporal convolutional network (TCN) and convolutional neural network (CNN), and BiLSTM with attention-based feature enhancement. First, SVMD is employed to adaptively decompose multi-sensor signals, and components with low Pearson correlation coefficients are removed as noise-dominated components. The remaining components are reconstructed to obtain denoised signals with improved quality. Subsequently, parallel TCN and CNN branches are constructed to extract temporal and spatial features, respectively, enabling comprehensive representation of spatiotemporal characteristics from multi-channel signals. Finally, a BiLSTM combined with an attention mechanism is utilized to model long-term dependencies and emphasize discriminative features for accurate fault classification. The proposed method is evaluated on a public PMSM dataset containing eight sensor channels and nine operating states. Experimental results demonstrate that the proposed framework achieves an accuracy of 98.5%, outperforming several existing representative models. Full article

Solar energy makes a considerable contribution to global power generation, necessitating photovoltaic (PV) fault detection to ensure high yields in solar power systems. In real-world solar installations, operational data are geographically dispersed, heterogeneous, and sensitive, which imposes privacy restrictions. Existing PV fault-detection techniques have relied on centralized training, requiring raw image data from multiple plants to be collected at a single server, which has led to privacy risks, bias from heterogeneous datasets, and scalability issues. To overcome these challenges, this research provides a novel FL-based PV fault-detection model called Federated Convolutional Transformer Network (Fed-CVTNet), which combines Convolutional Neural Network (CNN) and Vision Transformer (ViT) architecture within the FL framework. Initially, the raw images are pre-processed to enhance the input quality, and the Region of Interest (ROI) is identified via a pretrained YOLO model. Then, the proposed Fed-CVTNet facilitates networked learning among many geographically dispersed clients by only sharing model updates using federated averaging (FedAvg). The experimental findings illustrate that the proposed FL model shows substantial quality improvements compared to the customized CNN-ViT models trained on a dataset of 5600 images and the CNN-ViT models that lack federated aggregation. The highest accuracy achieved by the proposed technique is 98.99%; the proposed Fed-CVTNet has better sensitivity (98.15%) and specificity (99.21%), and much lower false positive and false negative rates than its centralized counterparts. The comparative analysis establishes that federated weight aggregation outperforms centralized baseline models by 2.3% and is effective in reducing the data privacy risk. Full article

Background: Artificial intelligence (AI) has emerged as a transformative technology across dermatological practice, from automated lesion classification to whole-slide pathology analysis. Despite rapid growth in primary studies, a comprehensive synthesis of diagnostic performance, application breadth, and real-world implementation remains lacking. Methods: We conducted a PRISMA systematic review and meta-analysis of studies published from January 2000 to March 2025. We searched the PubMed, Cochrane, and ScienceDirect databases for studies reporting AI diagnostic performance in dermatology. Results: Of 30 included studies (28 valid after exclusion of two retracted publications), 60% focused on melanoma and related lesions. AI diagnostic performance improved markedly over five identified temporal eras (2003–2025), with a pooled AUROC of 0.92 (95% CI 0.87–0.96), Reitsma sensitivity of 0.88 (0.82–0.93), and Reitsma specificity of 0.85 (0.75–0.91). AI matched or surpassed specialist dermatologists in 71% of direct comparisons. Three randomized controlled trials (RCTs) were identified, with heterogeneous findings across different clinical applications: AI assistance significantly improved non-expert diagnostic accuracy in one trial (53.9% vs. 43.8%; p = 0.019), significantly reduced acne severity via personalized treatment recommendations in a second, and showed non-inferior diagnostic performance, but was not cost-effective in the third. The sole cost-effectiveness analysis found AI-assisted surveillance not cost-effective over a 2-year horizon. Conclusions: AI achieves dermatologist-level diagnostic accuracy in controlled settings; however, real-world evidence, algorithmic equity across skin phototypes, and health economic viability remain critical unresolved challenges. Prospective validation, mandatory demographic subgroup reporting, and cost-effectiveness modeling are essential prerequisites for safe and equitable clinical implementation. Full article

Electric control valves (ECVs) are core control components in process industries such as petrochemicals and power generation, and their operational reliability directly affects system safety and energy efficiency. However, frequent changes in the working conditions of ECVs cause vibration signals to exhibit strong nonlinearity and non-stationarity, which leads to the loss of high-frequency transient features, difficulty in extracting weak faults, and cross-condition domain shifts. These issues severely limit the generalization ability of existing fault diagnosis methods. To address this, this study proposes a collaborative fault diagnosis framework that combines a miniaturized high-frequency data acquisition system with a multi-scale attention-conditioned domain adversarial network (MS-ACDAN). First, a miniaturized high-speed data acquisition system is developed based on a field-programmable gate array (FPGA) to enable lossless acquisition of high-frequency transient signals. Subsequently, the original vibration signals are decomposed, filtered, and reconstructed using Adaptive Noise-Complete Empirical Mode Decomposition (CEEMDAN) and the Comprehensive Sensitivity Index (CSI) to generate feature-enhanced signals with high signal-to-noise ratios. Next, a feature extractor combining a one-dimensional convolutional neural network (1D-CNN) with a channel attention mechanism is constructed to automatically focus on key fault frequency band features while suppressing redundant information. Finally, a Conditional Adversarial Network (CDAN) is introduced for transfer learning. By establishing a conditional dependency between class prediction and feature representation. This approach achieves domain alignment while preserving the discriminative features of the data, thereby overcoming the limitation of traditional domain adaptation methods that ignore category information. The experimental results show that the proposed fault diagnosis framework demonstrates high recognition performance in various transfer tasks. Furthermore, even under extreme industrial noise conditions of 0 dB, the framework exhibits good diagnostic robustness. This research provides a theoretical basis and technical solution for addressing the fault diagnosis of critical control equipment under complex and variable working conditions. Full article

Accurate characterization of picosecond broadband pulse signals is important for radio-frequency electronics, radar detection and high-speed communications. However, bandwidth-limited oscilloscopes attenuate high-frequency components, broaden rising edges, reduce peak amplitudes and introduce ringing distortion. Conventional regularized deconvolution relies heavily on manual parameter tuning and is vulnerable to noise, while Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) models suffer from long training time and insufficient capability to capture global temporal dependencies. Here we present a waveform reconstruction method integrating convolutional neural networks, Transformer and learnable physics-guided deconvolution constraints. The model simultaneously extracts local waveform details and models global temporal correlations, which improves training efficiency and noise robustness, and restores pulse rising edges, peak amplitudes and overall waveform morphology with higher precision. The proposed method offers a low-cost software solution to realize high-precision measurement of high-speed pulse signals without hardware upgrades. Full article

The precise extraction of field roads in hilly and terrace farmlands poses great challenges for remote sensing analysis. Farm roads are narrow and irregular, with spectra close to crop ridges, and frequently covered by tree and building shadows. Traditional segmentation methods easily produce missed detection, broken and distorted road outlines. Most existing deep learning networks are optimized for cities and lack generalization to complex rural scenes. This work presents a multi-scale feature enhancement and edge-aware network (BRPE-Unet), an improved U-Net for accurate rural road extraction from high-resolution remote sensing images. An optimized DenseASPP block is inserted into encoders to fix U-Net’s weak multi-scale feature capacity and fuse tiny road details with global context. Parallel Polarized Self-Attention is added to skip links to highlight road areas and suppress crop and shadow noise. A new Adaptive Edge Refinement module is placed at the decoder output, using image-dependent threshold generation and multi-directional gradient information to refine blurred and fragmented road boundaries and improve segmentation continuity. Tests on DeepGlobe Road and WHU-RuR+ datasets yield IoUs of 67.54% and 52.30%, respectively, and the proposed BRPE-UNet demonstrates competitive road extraction performance among the representative methods evaluated in this study. BRPE-UNet can reliably extract rural roads and offers technical support for farm modernization and rural revitalization. Full article

To mitigate the delay spread and inter-symbol interference (ISI) induced by optical scattering in underwater wireless optical communication (UWOC), this paper introduces long short-term memory (LSTM), a convolutional block attention module (CBAM), and residual connections into a convolutional neural network autoencoder (CNN-AE), and proposes a CNN-LSTM-AE-based end-to-end DC-biased optical orthogonal frequency division multiplexing (DCO-OFDM) system. In the proposed system, convolutional layers in the encoder serve to extract local features; CBAM adaptively weights salient features along the channel and spatial dimensions; LSTM layers model the temporal dependencies of signal sequences; and residual connections are incorporated to improve the learning capability for subtle signal features, thereby enhancing the robustness of the system against multipath channels. A symmetric structure is adopted at the receiver, ultimately enabling end-to-end signal recovery. Simulation results show that, under typical clear ocean and coastal ocean channel conditions, the proposed system outperforms end-to-end systems based on a fully connected autoencoder (FC-AE) and a CNN-AE at different modulation orders, i.e., different numbers of bits per symbol. For example, under strong scattering conditions in the coastal ocean channel, when the number of bits per symbol is 2 and the bit error rate (BER) is ${10}^{-3}$, the proposed system achieves signal-to-noise ratio (SNR) gains of approximately 3.33 dB and 1.82 dB over the two baselines. In terms of block error rate (BLER), SNR gains of approximately 4.52 dB and 2.19 dB are achieved over the two comparison systems, which substantiates the superior end-to-end transmission reliability of the proposed system. Full article

Electrical induction generators are vulnerable to winding faults that can degrade operational reliability and lead to unplanned maintenance. This study proposes a multiple-convolutional neural network (Multi-CNN) extreme ensemble learning framework for intelligent fault diagnosis and maintenance decision support using three-phase current and voltage measurements. Experimental recordings representing healthy operation, inter-turn faults, and inter-winding faults were segmented into non-overlapping 200-sample windows. Hjorth activity, mobility, and complexity were calculated for the three-phase current signals and the three-phase voltage signals, producing 18 features for each of 900 instances. Four convolutional neural network architectures were trained, and their class-probability outputs were combined through an extreme learning machine. Stratified blocked five-fold cross-validation was used to evaluate the models while preserving the chronological structure of the data. The proposed ensemble achieved 98.111% accuracy, 98.146% precision, 98.111% recall, 98.108% F1-score, and 97.167% Matthews correlation coefficient, correctly classifying 883 of 900 out-of-fold instances. It also attained a macro-averaged area under the receiver operating characteristic curve of 0.995. These results demonstrate that Hjorth-based electrical-signal characterization and Multi-CNN ensemble fusion can provide accurate and computationally efficient support for fault identification and predictive maintenance decisions in electrical induction generators. Full article

Reliable short-term freeway traffic forecasting is essential for proactive traffic management, including congestion warning, ramp metering support, and traveler information services. However, many existing forecasting models treat spatial dependencies as synchronous or weakly directional, limiting their ability to represent delayed upstream–downstream traffic propagation. This study proposes Traffic-DiMAGNet, a lightweight and interpretable lag-aware directed spatio-temporal graph neural network for real-time freeway flow forecasting. The model constructs a sparse directed sensor dependency graph by integrating physical road adjacency, training-set lead–lag traffic priors, and learnable source–target node embeddings. A lag-aware bidirectional propagation module then shifts inter-sensor messages according to estimated propagation delays, while directed random-walk normalization, directional gating, and multi-scale causal convolutions preserve asymmetric traffic semantics with low computational cost. Experiments on four Caltrans PeMS datasets show that Traffic-DiMAGNet consistently outperforms recurrent, diffusion-based, attention-based, and adaptive-graph baselines regarding MAE, RMSE, and MAPE values. The learned directed lag structures provide interpretable propagation information, and the lightweight architecture supports rolling 5 min forecasting, indicating practical potential for real-time freeway monitoring and proactive traffic management. Full article

Sparse count data, such as patent document–keyword matrices, often contain excessive zeros and overdispersion, making conventional distance-based clustering methods less suitable. This study proposes a zero-inflated likelihood-based graph neural clustering method with a zero-inflated negative binomial likelihood, denoted as ZIL-GNC-ZINB. The proposed method combines graph-smoothed node representations with cluster-specific zero-inflation probabilities and count-intensity parameters. By incorporating graph-neighborhood information into the cluster membership update, ZIL-GNC-ZINB jointly accounts for structural zeros, overdispersion, and local graph relationships among observations. The proposed method was applied to a quantum-computing patent document–term matrix consisting of 9416 patent documents and 82 reduced keywords. Compared with K-means, K-means clustering based on principal component analysis (PCA+K-means), and K-means clustering based on graph convolutional networks (GCN+K-means), ZIL-GNC-ZINB achieved the best performance in terms of negative log-likelihood (NLL), zero area underneath the receiver operating characteristic (ROC) curve (AUC), and zero Brier score. The resulting clusters revealed interpretable quantum-computing sub-technologies, including photonic qubit control, hybrid quantum–classical computing, superconducting qubit hardware, and quantum security networks. Simulation experiments under zero proportions of 0.5, 0.7, and 0.9 further showed that the proposed method becomes increasingly effective as zero inflation becomes more severe. In the extreme zero-inflation setting, ZIL-GNC-ZINB achieved the best performance in NLL, Zero AUC, adjusted rand index (ARI), normalized mutual information (NMI), and clustering accuracy (ACC). These results demonstrate that zero-inflated likelihood modeling combined with graph-based clustering provides an effective and interpretable framework for sparse high-dimensional count data. Full article

Background/Objectives: Artificial intelligence (AI) is rapidly transitioning from a theoretical concept to an active clinical tool in dentistry and medicine. These are fields in which the current market-driven shift towards automation is observed. The aim of this study was to describe the possible applications of artificial intelligence particularly in oral surgery and maxillofacial surgery, fields that are deeply rooted in manual work with patients. Methods: A review of current literature was conducted using PubMed/MEDLINE, Scopus, and Web of Science databases. Keywords like: “oral surgery”, “maxillofacial surgery”, “implantology”, “dentistry”, “artificial intelligence”, and “orthognathic surgery” and their combinations were applied. Non-English articles were excluded. Results: Based on the literature, current applications of artificial intelligence in oral surgery and maxillofacial surgery are presented. Results indicate high efficacy of convolutional neural networks (CNNs) in radiographic triage and large language models (LLMs) in postoperative patient communication. Conclusions: Artificial intelligence is a useful tool that can significantly improve the work of physicians. However, it should not be considered a “replacement” for physicians, especially in fields requiring manual work. It can provide important guidance for patients and be a form of support for physicians. Full article