Search Results (4,839)

Major depressive disorder (MDD) is a prevalent mental health condition that requires accurate and objective diagnostic tools. Electroencephalogram (EEG) signals provide valuable insights into brain activity and have been widely studied for mental disorder classification. In this study, we propose a novel DML + ANFIS framework that integrates deep metric learning (DML) with an adaptive neuro-fuzzy inference system (ANFIS) for the automated diagnosis of major depressive disorder (MDD) using EEG time series signals. Time–frequency features are first extracted from raw EEG signals using the short-time Fourier transform (STFT) and the continuous wavelet transform (CWT). These features are then embedded into a low-dimensional space using a DML approach, which enhances the inter-class separability between MDD and healthy control (HC) groups in the feature space. The resulting time–frequency feature embeddings are finally classified using an ANFIS, which integrates fuzzy logic-based nonlinear inference with deep metric learning. The proposed DML + ANFIS framework was evaluated on a publicly available EEG dataset comprising MDD patients and healthy control (HC) subjects. Under subject-dependent evaluation, the STFT-based DML + ANFIS and CWT-based models achieved an accuracy of 92.07% and 98.41% and an AUC of 97.28% and 99.50%, respectively. Additional experiments using subject-independent cross-validation demonstrated reduced but consistent performance trends, thus indicating the framework’s ability to generalize to unseen subjects. Comparative experiments showed that the proposed approach generally outperformed conventional deep learning models, including Bi-LSTM, 2D CNN, and DML + NN, under identical experimental conditions. Notably, the DML module compressed 1280-dimensional EEG features into a 10-dimensional embedding, thus achieving substantial dimensionality reduction while preserving discriminative information. These results suggest that the proposed DML + ANFIS framework provides an effective balance between classification performance, generalization capability, and computational efficiency for EEG-based MDD diagnosis. Full article

Background: Seronegative rheumatoid arthritis (SNRA), defined by the absence of rheumatoid factor (RF) and anti-citrullinated peptide antibodies (ACPA), represents 20–30% of rheumatoid arthritis cases. Once considered a milder phenotype, SNRA is now recognised as a heterogeneous entity in which a substantial subset of patients develops structural progression comparable to seropositive RA. The binary RF/ACPA-based definition is increasingly viewed as insufficient, as the broader anti-modified protein antibody (AMPA) family—including antibodies against carbamylated, acetylated and malondialdehyde–acetaldehyde–modified proteins—indicates that many “seronegative” patients may harbour unconventional humoral autoimmunity undetected by standard assays. Objectives: To synthesise contemporary insights into the epidemiology, immunopathology, diagnostic challenges and therapeutic management of SNRA, with emphasis on erosive versus non-erosive phenotypes and the implications of the AMPA paradigm. Methods: A comprehensive literature search of PubMed, Cochrane Library and Google Scholar identified randomised trials, observational cohorts and systematic reviews, with focus on studies published within the past decade. Results: SNRA displays partially distinct immune features, including lower formation of tertiary lymphoid structures and variable activation of innate inflammatory circuits. However, the traditional adaptive–versus–innate dichotomy is overly reductionist. Growing evidence suggests that unconventional humoral responses directed against non-classical post-translational modifications may be present in a proportion of RF/ACPA-negative patients. Additional qualitative dimensions—such as IgA isotypes and fine-specificity profiles—represent further heterogeneity with potential prognostic significance. Although ACPA remains the strongest predictor of erosive progression, up to one-third of seronegative patients develop erosions within five years. The 2010 ACR/EULAR criteria may delay diagnosis in SNRA. Cytokine inhibitors and JAK inhibitors show largely serostatus-independent efficacy, whereas B-cell and T-cell–targeted therapies demonstrate attenuated responses in SNRA. Conclusions: SNRA is clinically and immunologically diverse. Integrating the AMPA framework is essential for refining classification and prognostication. Distinguishing erosive from non-erosive forms may guide treatment, while future work should prioritise biomarkers predicting progression and therapeutic response. Full article

Deep learning for semantic segmentation has made significant advances in recent years, achieving state-of-the-art performance. Medical image segmentation, as a key component of healthcare systems, plays a vital role in the diagnosis and treatment planning of diseases. Due to the fractal and scale-invariant nature of biological structures, effective medical image segmentation requires models capable of capturing hierarchical and self-similar representations across multiple spatial scales. In this paper, a Recurrent Neural Network (RNN) is explored within the Convolutional Neural Network (CNN) and Vision Transformer (ViT)-based hybrid U-shape network, named RCV-UNet. First, the ViT-based layer was developed in the bottleneck to effectively capture the global context of an image and establish long-range dependencies through the self-attention mechanism. Second, recurrent residual convolutional blocks (RRCBs) were introduced in both the encoder and decoder to enhance the ability to capture local features and preserve fine details. Third, by integrating the global feature extraction capability of ViT with the local feature enhancement strength of RRCBs, RCV-UNet achieved promising global consistency and boundary refinement, addressing key challenges in medical image segmentation. From a fractal–fractional perspective, the multi-scale encoder–decoder hierarchy and attention-driven aggregation in RCV-UNet naturally accommodate fractal-like, scale-invariant regularity, while the recurrent and residual connections approximate fractional-order dynamics in feature propagation, enabling continuous and memory-aware representation learning. The proposed RCV-UNet was evaluated on four different modalities of images, including CT, MRI, Dermoscopy, and ultrasound, using the Synapse, ACDC, ISIC 2018, and BUSI datasets. Experimental results demonstrate that RCV-UNet outperforms other popular baseline methods, achieving strong performance across different segmentation tasks. The code of the proposed method will be made publicly available. Full article

Background: Breast cancer remains a leading cause of mortality among women worldwide, underscoring the need for timely and accurate detection. Conventional mammographic diagnosis, while widely used, is limited by subjectivity and variability in interpretation. Recent advances in deep learning (DL) have improved automated detection; however, the black-box nature of these models raises concerns regarding clinical trust and interpretability. Methods: To address this, we propose an explainable DL framework for breast cancer classification using mammographic images. The approach employs contrast limited adaptive histogram equalization (CLAHE)-based preprocessing to enhance lesion contrast, EfficientNetV2 for feature extraction, and the convolutional block attention module (CBAM) to refine salient features. For interpretability, gradient-weighted class activation mapping++ (Grad-CAM++) is used to highlight discriminative regions influencing predictions. Results: The framework is evaluated on three publicly available datasets—MIAS, DDSM, and InBreast—individually and under cross-dataset settings. Results demonstrate superior performance over existing methods, achieving classification accuracies of 99.85%, 99.40%, and 99.70% on MIAS, DDSM, and InBreast, respectively, with corresponding F1-scores of 99.75%, 99.10%, and 99.55%. Confusion matrix analysis confirms excellent sensitivity for malignant cases, and time complexity assessments show reduced training and inference overhead compared to conventional deep models. Conclusions: The framework thus provides a robust and interpretable solution for mammogram-based breast cancer screening. Full article

Photoacoustic imaging (PAI) integrates the high-contrast merits of optical imaging with the high-spatial-resolution advantages of acoustic imaging, enabling the acquisition of three-dimensional images with deep tissue penetration (up to several centimeters) for in vivo disease detection and diagnosis. Among various photoacoustic nanoagents, gold nanomaterials (GNMs) have been widely explored for the PAI-based imaging analysis and photothermal therapy of diseases, owing to their strong near-infrared (NIR) absorption, which can generate distinct photoacoustic signals in deep tissues. This review focuses on recent advances and achievements in the development of functionalized gold nanoprobes, including Janus gold nanoprobes, gold nanocomposite probes (such as functionally coated GNMs and GNMs-loaded nanocarriers), and gold nanoaggregate probes (e.g., pre-assembly of GNMs and in situ aggregation of GNMs). The multifunctionalization of GNMs can enhance their PAI performance by shifting absorption to the NIR-I and NIR-II regions, while simultaneously imparting additional functionalities such as targeted delivery to disease sites and specific responsiveness to disease biomarkers. These features can render functionalized GNMs-based nanoprobes highly suitable for PAI-based analysis and the precise detection of various pathological conditions, including bacterial infections, tumors, kidney injury, and disorders affecting the ocular, gastrointestinal, cardiovascular, visceral, and lymphatic systems. Finally, this review provides a concise summary of biosafety evaluation and outlines the current challenges and future perspectives in optimizing the GNMs-based PAI methods, highlighting their potential to enhance the rapid and precise diagnosis of diseases in the future. Full article

Background: Optical coherence tomography (OCT) can quantify the morphology and dimensions of a macular hole for diagnosis and treatment planning. Objective: The aim of this study was to perform automatic segmentation of macular holes (MHs) and cysts from OCT macular volumes using a deep learning-based model and to quantitatively evaluate decision reliability using the model’s focus regions and GradientSHAP-based explainability. Methods: In this study, we automatically segmented MHs and cysts in OCT images from the open-access OIMHS dataset. The dataset comprises 125 eyes from 119 patients and 3859 OCT B-scans. OCT B-scan slices were input to a UNet-48-based model with a 2.5D stacking strategy. Performance was evaluated using Dice and intersection-over-union (IoU), boundary accuracy was evaluated using the 95th-percentile Hausdorff distance (HD95), and calibration was evaluated using the expected calibration error (ECE). Explainability was quantified from GradientSHAP maps using lesion coverage and spatial focus metrics: Attribution Precision in Lesion (APILτ), which is the proportion of attributions (SHAP contributions) falling inside the lesion; Attribution Recall in Lesion (ARILτ), which is the proportion of the true lesion covered by the attributions; and leakage (Leakτ = 1 − APILτ), which is the proportion of attributions falling outside the lesion. Spatial focus was monitored using the center-of-mass distance (COM-dist), which is the Euclidean distance between the attribution center and the segmentation center. All metrics were calculated using the top τ% of the pixels with the highest SHAP values. SHAP features were clustered using PCA and k-means. Explanations were calculated using the clinical mask in ground truth (GT) mode and the model segmentation in prediction (Pred) mode. Results: The Dice/IoU values for holes and cysts were 0.94/0.91 and 0.87/0.81, respectively. Across lesion classes, HD95 = 6 px and ECE = 0.008, indicating good boundary accuracy and calibration. In GT mode (τ = 20), three regimes were observed: (i) retina-dominant: high ARIL (hole: 0.659; cyst: 0.654), high Leak (hole: 0.983; cyst: 0.988), and low COM-dist (hole: 7.84 px; cyst: 6.91 px), with the focus lying within the retina and largely confined to the retinal tissue; (ii) peri-lesional: highest ARIL (hole: 0.684; cyst: 0.719), relatively lower Leak (hole: 0.917; cyst: 0.940), and medium/high COM-dist (hole: 16.22 px; cyst: 10.17 px), with the focus located around the lesion; (iii) narrow-coverage: primarily seen for cysts in GT mode (ARIL: 0.494; Leak: 1.000; COM-dist: 52.02 px), with markedly reduced coverage. In Pred mode, the ARIL20 for holes increased in the retina-dominant cluster (0.758) and COM-dist decreased (6.24 px), indicating better agreement with the model segmentation. Conclusions: The model exhibited high accuracy and good calibration for MH and cyst segmentation in OCT images. Quantitative characterization of SHAP validated the model results. In the clinic, peri-lesion and narrow-coverage conditions are the key situations that require careful interpretation. Full article

We propose a novel continual self-supervised learning (CSSL) framework for simultaneously learning diverse features from multi-window-obtained chest computed tomography (CT) images and ensuring data privacy. Achieving a robust and highly generalizable model in medical image diagnosis is challenging, mainly because of issues, such as the scarcity of large-scale, accurately annotated datasets and domain shifts inherent to dynamic healthcare environments. Specifically, in chest CT, these domain shifts often arise from differences in window settings, which are optimized for distinct clinical purposes. Previous CSSL frameworks often mitigated domain shift by reusing past data, a typically impractical approach owing to privacy constraints. Our approach addresses these challenges by effectively capturing the relationship between previously learned knowledge and new information across different training stages through continual pretraining on unlabeled images. Specifically, by incorporating a latent replay-based mechanism into CSSL, our method mitigates catastrophic forgetting due to domain shifts during continual pretraining while ensuring data privacy. Additionally, we introduce a feature distillation technique that integrates Wasserstein distance-based knowledge distillation and batch-knowledge ensemble, enhancing the ability of the model to learn meaningful, domain-shift-robust representations. Finally, we validate our approach using chest CT images obtained across two different window settings, demonstrating superior performance compared with other approaches. Full article

Background/Objectives: Alzheimer’s disease is the leading cause of dementia, marked by progressive cognitive decline and a severe socioeconomic burden. Early and accurate diagnosis is crucial to enhancing patient outcomes, yet traditional clinical and imaging assessments are often limited in sensitivity, particularly at early stages. This study presents a dual-modal framework that integrates symptom-based clinical data with magnetic resonance imaging (MRI) using machine learning (ML) and deep learning (DL) models, enhanced by explainable AI (XAI). Methods: Four ML classifiers—K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Decision Tree (DT), and Random Forest (RF)—were trained on demographic and clinical features. For stage-wise classification, five DL models—CNN, EfficientNetB3, DenseNet-121, ResNet-50, and MobileNetV2—were applied to MRI scans. Interpretability was incorporated through SHAP and Grad-CAM visualizations. Results: Random Forest achieves the highest accuracy of 97% on clinical data, while CNN achieves the best overall performance of 94% in MRI-based staging. SHAP and Grad-CAM were used to find clinically relevant characteristics and brain areas, including hippocampal atrophy and ventricular enlargement. Conclusions: Integrating clinical and imaging data and interpretable AI improves the accuracy and reliability of AD staging. The proposed model offers a valid and clear diagnostic route, which can assist clinicians in making timely diagnoses and adjusting individual treatment. Full article

Background/Objectives: Cardiovascular diseases are the leading cause of mortality worldwide according to the World Health Organization (WHO), highlighting the need for accessible tools for early detection. Automated classification systems based on signal processing and machine learning offer a non-invasive alternative to support clinical diagnosis. Methods: This study implements and evaluates machine learning models for distinguishing normal and abnormal heart sounds using a hybrid feature extraction approach. Recordings labeled as normal, murmur, and extrasystolic were obtained from the PASCAL dataset and subsequently binarized into two classes. Multiple numerical datasets were generated through statistical features derived from Mel-Frequency Cepstral Coefficients (MFCCs) and Daubechies wavelet analysis. Each dataset was standardized and used to train four classifiers: support vector machines, logistic regression, random forests, and decision trees. Results: Model performance was assessed using accuracy, precision, recall, specificity, F1-score, and area under curve. All classifiers achieved notable results; however, the support vector machine model trained with 26 MFCCs and Daubechies-4 wavelet coefficients obtained the best performance. Conclusions: These findings demonstrate that the proposed hybrid MFCC–Wavelet framework provides competitive diagnostic accuracy and represents a lightweight, interpretable, and computationally efficient solution for computer-aided auscultation and early cardiovascular screening. Full article

Background/Objectives: Tyrosinemia type 1 (HT-1) is a treatable inherited disorder characterized by disrupted tyrosine metabolism, leading to severe liver, renal, and occasionally neurological dysfunction. Early diagnosis by newborn screening markedly reduces the risk of hepatocellular carcinoma (HCC), the most serious complication. A deeper understanding of HT-1 pathophysiology is necessary to prevent disease complications and improve diagnostic and therapeutic strategies. This study explored the untargeted serum metabolomic profiles of HT-1 patients. Methods: High-resolution untargeted metabolomics coupled with liquid chromatography was applied for serum analysis of 16 late-diagnosed Chilean HT-1 patients on nitisinone (NTBC) therapy and 16 age- and sex-matched controls. The statistically significant up- and down-regulated features were used for annotation and association with different metabolic pathways. Results: Untargeted metabolomics revealed 1066 features significantly changed in HT-1 patients. Increased metabolites included aromatic compounds, medium- and long-chain acyl-carnitines, bile acids (prevalently taurine-conjugated), indole-based compounds, modified nucleosides and nucleobases. Decreased metabolites were mainly related to lipid class, including lysophosphatidylcholines, lysophosphatidic acids, long-chain fatty acids, and acylglycerols. Conclusions: Untargeted metabolomics showed perturbation of tyrosine- and tryptophan-related pathways and described a novel HT-1 metabolomic pattern demonstrating net dysregulation of lipid and bile acid metabolism in NTBC-treated patients with delay diagnoses. Increased acylcarnitines, taurine-conjugated bile acids, modified nucleobases, and reduced lysophosphatidylcholines overlap with the metabolomic pattern previously reported in Wnt/β-catenin-associated HCC. Although direct mechanistic link cannot be established in this study, these alterations may reflect persistent disease-related metabolic adaptations and warrant further investigation to clarify their potential relevance with long-term complications. Full article

Flywheel energy storage systems play an important role in frequency regulation and power quality control within modern power grids, yet the fault signals generated by defects in their rolling bearings are typically indistinct, making direct diagnosis difficult. Raw noisy signals often yield unsatisfactory diagnostic performance when directly processed by neural networks. Although MOMEDA (Multipoint Optimal Minimum Entropy Deconvolution Adjusted) can effectively extract impulsive fault components, its performance is highly dependent on the selected fault period and filter length. To address these issues, this paper proposes an improved fault diagnosis method that integrates MOMEDA-based periodic extraction with a neural network classifier. The Artificial Fish Swarm Algorithm (AFSA) is employed to adaptively determine the key parameters of MOMEDA using multi-point kurtosis as the optimization objective, and the optimized parameters are used to enhance impulsive fault features. The filtered signals are then converted into image representations and fed into a ResNet-18 network (a compact 18-layer deep convolutional neural network from the residual network family) to achieve intelligent identification and classification of bearing faults. Experimental results demonstrate that the proposed method can effectively extract and diagnose bearing fault signals. Full article

Drivetrain systems operate under varying operating conditions (OCs), which often obscure early-stage fault signatures and hinder robust condition monitoring (CM). This work introduces an AI digital platform developed during the EEDRIVEN project, featuring a holistic CM framework that integrates statistical time series methods—using Generalized AutoRegressive (GAR) models in a multiple model fault diagnosis scheme—with deep learning approaches, including autoencoders and convolutional neural networks, enhanced through a dedicated decision fusion methodology. The platform addresses all key CM tasks, including fault detection, fault type identification, fault severity characterization, and remaining useful life (RUL) estimation, which is performed using a dynamics-informed health indicator derived from GAR parameters and a simple linear Wiener process model. Training for the platform relies on a limited set of experimental vibration signals from the physical drivetrain, augmented with high-fidelity multibody dynamics simulations and surrogate-model realizations to ensure coverage of the full space of OCs and fault scenarios. Its performance is validated on hundreds of inspection experiments using confusion matrices, ROC curves, and metric-based plots, while the decision fusion scheme significantly strengthens diagnostic reliability across the CM stages. The results demonstrate near-perfect fault detection (99.8%), 97.8% accuracy in fault type identification, and over 96% in severity characterization. Moreover, the method yields reliable early-stage RUL estimates for the outer gear of the drivetrain, with normalized errors < 20% and consistently narrow confidence bounds, which confirms the platform’s robustness and practicality for real-world drivetrain systems monitoring. Full article

Electroencephalography (EEG) has emerged as a powerful, non-invasive modality for investigating neurological and oculomotor disorders, particularly when combined with advances in artificial intelligence (AI). This systematic review examines recent progress in machine learning (ML) and deep learning (DL) techniques applied to EEG-based analysis for the diagnosis, classification, and monitoring of neurological conditions, including oculomotor-related disorders. Following the PRISMA guidelines, a structured literature search was conducted across major scientific databases, resulting in the inclusion of 15 peer-reviewed studies published over the last decade. The reviewed works encompass a range of neurological and ocular-related disorders and employ diverse AI models, from conventional ML algorithms to advanced DL architectures capable of learning complex spatiotemporal representations of neural signals. Key trends in feature extraction, signal representation, model design, and validation strategies are synthesized here to highlight methodological advancements and common challenges. While the reviewed studies demonstrate the growing potential of AI-enhanced EEG analysis for supporting clinical decision-making, limitations such as small sample sizes, heterogeneous datasets, and limited external validation remain prevalent. Addressing these challenges through standardized methodologies, larger multi-center datasets, and robust validation frameworks will be essential for translating EEG-driven AI approaches into reliable clinical applications. Overall, this review provides a comprehensive overview of current methodologies and future directions for AI-driven EEG analysis in neurological and oculomotor disorder assessment. Full article

Gears, as critical components of rotating machinery, are prone to wear and fracture due to their complex structural dynamics and harsh operating conditions, leading to catastrophic failures, economic losses, and safety risks. AE technology enables real-time fault diagnosis by capturing stress wave emissions from material defects with high sensitivity. However, mechanical background noise significantly corrupts AE signals, while optimal selection of gear health indicators remains challenging, critically impacting fault feature extraction accuracy. This study develops an adaptive feature extraction method for fault diagnosis using AE. Through gear fault simulation experiments, VMD analyzes mode number and penalty factor effects on signal decomposition. Correlation coefficient-based reconstruction optimization is implemented. For feature selection challenges, SVM-RFE enables adaptive parameter ranking. Finally, SVM with optimized kernel parameters achieves effective fault classification. Optimized VMD enhances signal decomposition, while SVM-RFE reduces feature dimensionality, addressing manual selection uncertainty and computational redundancy. Experimental results demonstrate superior accuracy in gear fault classification. This study proposes an AE-based adaptive feature extraction method with three innovations: (1) establishing VMD parameter–decomposition quality relationships; (2) developing an SVM-RFE feature selection framework; (3) achieving high-accuracy gear fault classification. The method provides a novel technical approach for rotating machinery diagnostics with significant engineering value. Full article

Deep learning has become a popular topic among scholars and has attracted widespread attention. However, deep learning methods typically require large datasets to determine model parameters and can only process data in batches. To address the challenges of deep learning models, which rely on batch data and struggle to adapt to industrial streaming data scenarios in gearbox fault diagnosis, this study proposes an online gearbox fault diagnosis method based on a DeepWalk graph embedding-enhanced extreme learning machine (ELM) approach. The method constructs a graph structure in real time for each newly collected vibration signal, uses DeepWalk for unsupervised embedding learning, and extracts low-dimensional features with strong discriminative power. These features are then input into the ELM classifier to achieve adaptive fault type recognition and online incremental model updates. This method does not require historical data to be retrained, thus effectively overcoming the bottleneck of batch retraining and significantly improving diagnostic efficiency and resource utilization. The experimental results show that, under various operating conditions, the proposed method achieves fast and accurate diagnosis of multiple gearbox fault types, with an average accuracy consistently above 95%, thereby demonstrating excellent engineering applicability and real-time performance. Full article