Search Results (3,479)

Early diagnosis of Diabetic Retinopathy (DR) is critical for preventing irreversible vision loss, but precise lesion annotation by ophthalmologists is the dominant cost in building any clinical-grade DR detection model. The structural problem in real hospital settings is not labeling cost per se, but expert availability: ophthalmologists’ time is bounded by clinical duties, so the active-learning (AL) cycle can iterate only a handful of times in practice. We frame this constraint explicitly and ask which AL designs work best under a tight expert budget. We propose Virtuous Cycle, a Human-in-the-Loop (HITL) pipeline that integrates (i) a YOLOv8x-based object detector for microaneurysms, hemorrhages, and exudates, (ii) four AL sampling strategies (Average Confidence, Random, Hybrid-Diversity, Monte Carlo Dropout), and (iii) an in-hospital annotation platform (Diavision Studio) in which clinicians refine AI pre-labels rather than draw from scratch. We evaluate Virtuous Cycle on a real-world fundus dataset from the National Medical Center (NMC) across eight AL rounds, expanding the labeled pool from 81 images (R0) to 481 images (R8) within the actual expert-time budget of two ophthalmologists. Across three independent random seeds, random sampling dominates at cold start (mean mAP@50 $0.14\to 0.25$ over R0–R1), whereas Hybrid-Diversity converges to the highest mAP@50, Precision, and Recall by R7 (431 images; mAP@50 $0.40$, Precision $0.55$, Recall $0.41$), with MC Dropout close behind; by R8, the labeled pool is exhausted and all strategies converge to the same final model. A clinician crossover analysis of 36 paired clinical images, controlling for per-clinician speed bias and per-image difficulty bias, shows no statistically significant difference in overall per-image labeling time between AI-assisted and manual annotation ($p=0.52$), but a statistically significant increase in confirmed lesion detections under AI assistance ($p=0.0058$), driven predominantly (84–$100\%$ of the net increase) by microaneurysms, the lesion type most prone to being missed unaided. The results indicate that, under expert-budget constraints, AL strategy choice should be staged: random sampling for cold start, uncertainty-and-diversity sampling once the model has matured, and that AI assistance trades a modest, lesion-burden-dependent time cost for a measurable gain in the sensitivity of microaneurysm detection. Full article

Background: Lung cancer remains the leading cause of cancer-related mortality worldwide, and early diagnosis and accurate disease stratification are still major clinical challenges. Radiomics has emerged as a quantitative imaging approach that extracts high-dimensional features from radiological imaging, with applications in diagnosis, prognosis, radio genomics, and assessment of treatment response. However, its clinical translation is still limited by methodological heterogeneity and a lack of standardization. Aim: This narrative review synthesizes evidence from systematic reviews and meta-analyses on radiomics in thoracic imaging for lung cancer, focusing on clinical applications, methodological limitations, and translational challenges. Methods: A structured search was conducted in PubMed and Scopus using predefined keywords related to radiomics, lung cancer, and imaging modalities. Only peer-reviewed systematic reviews and meta-analyses published in English were included. In total, 27 studies were selected and synthesized using a structured narrative approach guided by the ANDJ checklist. A differential integrative framework was adopted to connect evidence from systematic reviews and meta-analyses with primary empirical studies and policy documents through an intermediate layer of translational recommendations, ensuring a multi-level and interpretation-driven synthesis. Results: Radiomics demonstrated consistent potential across multiple clinical domains, including lesion classification, histological differentiation, molecular profiling, prognostic stratification, and prediction of treatment response. Machine learning and deep learning approaches frequently improved predictive performance. However, key limitations were identified, including heterogeneity in imaging protocols, lack of external validation, small single-centre datasets, and limited reproducibility of radiomic features. Conclusions: Radiomics in lung cancer imaging shows strong clinical potential but remains constrained by methodological and translational barriers. Future progress will depend on standardization, external validation, multimodal data integration, and improved interpretability, alongside alignment with regulatory and clinical implementation frameworks. Full article

Early-stage maize leaf diseases and pests are difficult to detect due to their small size, low contrast, and complex backgrounds. AdaptiveLeaf is a lightweight multi-scale framework designed to improve the detection of such small targets. It integrates an Adaptive Kernel Lightweight Block (AKL-Block) for dynamic multi-scale feature extraction, a Feature Decomposition and Reconstruction (FDR) module to recover fine details such as lesion edges and spore clusters, and a Scale-Aware Gradient Boosting Loss (SAGB-Loss) to increase the training contribution of small targets. Experiments on 10,324 field-collected maize leaf images across eight disease and pest categories show that AdaptiveLeaf achieves a mean mAP@0.5 of 75.0% over three repeated runs and increases small-target AP from 28.1% to 32.8%, using only 2.52 M parameters and 5.3 GFLOPs. The framework balances accuracy and efficiency, making it suitable for real-time field inspection and precision agriculture. Full article

Background: Diphtheria is an acute infectious disease caused by Corynebacterium diphtheriae. Despite several worldwide outbreaks, it is now considered a rare disease by industrialized countries. Clinical manifestations usually account for oropharyngeal lesions, but rare cases of systemic involvement (mainly endocarditis) have been described among non-toxigenic strains. Case description: We report the case of a patient who experienced septic shock, disseminated intravascular coagulation and multiorgan failure due to Corynebacterium diphtheriae infection. The pathogen was further characterized as a highly toxigenic strain. Infective endocarditis with mitral and aortic valve vegetations led to early multiorgan septic embolization. Major stroke, liver function impairment, heart failure and acute kidney injury were the main findings. Unlike the typical forms of infection caused by this pathogen, there was no evidence of airway or skin involvement. Furthermore, apart from hemocultures, none of the other investigations (pharyngeal swabs, bronchoalveolar lavages, urine culture) ever tested positive for the bacteria. Conclusions: The report we present describes a case of C. diphtheriae infection with many atypical characteristics: (i) lack of any pathognomonic signs or symptoms; (ii) extensive endocarditic process (very uncommon for toxigenic strains); (iii) early septic emboli development, with rapid evolution to multiorgan failure; (iv) detection of disseminated intravascular coagulation. Despite disseminated intravascular coagulation being a known complication of septic shock, regardless of the etiological agent, according to our literature research, this is the second known case driven by C. diphtheriae infection in an adult. Full article

Epidermolysis bullosa (EB) comprises a group of rare inherited genodermatoses characterized by fragility and blistering of the skin and mucous membranes, chronic wounding, and significant morbidity including increased risk of squamous cell carcinoma in severe subtypes. Key unmet priorities include reducing diagnostic latency, establishing objective wound monitoring, enabling early detection of malignant transformation within chronic ulcerations, and developing therapies that durably modify disease progression. Artificial intelligence (AI) encompassing machine learning (ML), and deep learning (DL) is increasingly integrated into EB research and clinical practice to address these unmet needs. This structured narrative review synthesises current evidence on AI applications in EB spanning genetic diagnostics, wound assessment, inflammatory endotyping, drug repurposing, and emerging therapeutic technologies, and integrates evidence from registered clinical trials. In genomics, DL-based splicing prediction models and variant prioritisation frameworks accelerate pathogenic variant detection and reduce diagnostic latency. In wound care, convolutional neural networks-based platforms enable automated lesion segmentation and remote monitoring, while multimodal AI models predict healing trajectories and support stratification of wounds by chronicity. Computational transcriptomic analyses have identified candidate repurposing agents by reversing pathogenic gene expression signatures in EB tissue. Emerging convergence of AI with biosensors-integrated wound dressings and three-dimensional bioprinting of genetically corrected skin substitutes represents a transformative future direction. Translational barriers include limited EB-specific training datasets, algorithmic bias across diverse skin phototypes, the interpretability deficit of DL systems, and evolving regulatory frameworks for AI as a medical device. Expansion of internationally interoperable EB disease registries with standardised wound imaging protocols is identified as the single most impactful intervention to accelerate AI adoption. A minimum endpoint set for AI-assisted EB wound assessment, incorporating wound area trajectory, wound type classification, tissue composition, and paired patient-reported pain and itch scores, is proposed to standardise outcome reporting across future studies. Full article

Porcine reproductive and respiratory syndrome virus (PRRSV) is a major cause of disease in swine production, resulting in a high economic impact that has been exacerbated by the recent North American outbreaks of highly virulent PRRSV-2 strain (L1C.5). However, there is limited knowledge about how underlying systemic infections, particularly with this emergent PRRSV strain, affect calcium regulation and bone in pigs fed varying levels of vitamin D. To address this, the goal of this study was to determine the effect of different dietary vitamin D levels and PRRSV infection on calcium regulation and phenotypic changes in bone. Three-week-old pigs (n = 42) were assigned to four treatment groups: marginal dietary vitamin D3 (200 IU/kg) + PRRSV (n = 12), industry standard dietary vitamin D3 (1500 IU/kg) + PRRSV (n = 12), industry standard dietary vitamin D3 (1500 IU/kg) supplemented with 25-hydroxy-vitamin D3 (25-OH D3) (2000 IU/kg) + PRRSV (n = 12), or marginal dietary vitamin D3 without PRRSV inoculation (200 IU/kg) (control, n = 6). Following 26 days of dietary acclimation, assigned treatment groups were inoculated with a PRRSV-2 L1C.5 isolate. Blood samples were collected to evaluate the ionized calcium, 25-OH D3, calcium, phosphorus, and parathyroid hormone levels. The 2nd and 10th ribs were collected at 14 days post challenge for bone ash and density analysis, as well as examination of microscopic changes and scoring of the physis. High mortality was noted in all pigs infected with the virus, regardless of the vitamin D diet. Additionally, a significant depletion of serum calcium was observed at 7 DPC in infected animals, suggesting a high calcium demand at early stages of PRRSV infection. No significant differences in serum calcium, phosphorus, or ionized calcium concentrations were observed between dietary groups during the first 14 days of PRRSV-2 L1C infection. In pigs that succumbed to PRRSV at the early stage of infection, microscopic lesions of multifocal myelonecrosis were noted. This study provides the first report of microscopic changes of bone marrow necrosis and inflammation associated with PRRSV infection and demonstrates calcium dysregulation at the early stage of infection by this highly virulent PRRSV strain. Full article

Colitis-associated cancer (CAC) represents a major long-term complication in patients with ulcerative colitis (UC) and Crohn’s disease (CD), the two main forms of inflammatory bowel disease (IBD). Unlike sporadic colorectal cancer, CAC develops through a distinct inflammation–dysplasia–carcinoma sequence driven by chronic inflammation and complex molecular alterations. Early detection of dysplasia and accurate risk stratification remain critical challenges in IBD management. Conventional surveillance strategies, including endoscopy, histopathology, and immunohistochemistry, are time-consuming, operator-dependent, and may fail to identify early neoplastic changes. In this context, artificial intelligence (AI), including machine learning (ML) and deep learning (DL), has emerged as a promising approach to improve lesion detection, molecular characterization, and predictive risk modeling. Early studies, though limited, suggest that AI-based approaches may enhance the identification of dysplasia and CAC, improve risk prediction, and support personalized surveillance strategies. Furthermore, the integration of multimodal data, including clinical, endoscopic, histological, and molecular features, may further improve predictive performance and enable precision medicine approaches in IBD. This review summarizes current evidence on AI and ML applications for CAC detection and risk prediction in IBD, discusses technical and clinical challenges, and highlights future directions for integrating AI into routine clinical practice to improve surveillance and clinical outcomes in patients with IBD. Full article

Background: Noonan syndrome is a rare genetic disorder from the group of RASopathies, characterized by facial dysmorphism, congenital heart defects, hematologic abnormalities, and growth impairment. Case Presentation: We report the case of an 8-year-old girl with Noonan syndrome admitted for evaluation of abdominal pain and failure to thrive. Hematological evaluation before EGD did not identify contraindications to biopsy, and initial laboratory tests, including coagulation parameters, were normal. Several hours after upper gastrointestinal endoscopy, the patient developed abdominal pain and coffee-ground vomiting. Abdominal ultrasonography revealed an intramural duodenal hematoma (58 × 37 mm), which was confirmed and further characterized by computed tomography as an extensive, long-segment lesion involving the duodenum. Progressive anemia required transfusion of blood products. Conservative management, including nasogastric decompression, parenteral nutrition, and pharmacological treatment, was implemented. Despite the severity and prolonged clinical course, gradual clinical and radiological improvement was achieved, and the patient was discharged in good general condition after one month. Conclusions: Intramural duodenal hematoma is an extremely rare complication of upper gastrointestinal endoscopy with duodenal biopsy. This case highlights the importance of individualized assessment and close monitoring in patients with Noonan syndrome, and indicates that this complication should be considered early when abdominal pain, vomiting, or progressive anemia develops after the procedure, even when hematological evaluation and baseline coagulation parameters are reassuring. Full article

Background/Objectives: Oral cavity squamous cell carcinoma (OSCC) remains a major global health burden, with outcomes strongly dependent on stage at diagnosis. Although the oral cavity is directly accessible to clinical examination, many cases are still detected at advanced stages. This narrative review aimed to summarize current trends in OSCC diagnosis and early monitoring, with emphasis on non-invasive and minimally invasive techniques and biomarkers. Methods: A semi-systematic narrative literature search was conducted in PubMed/MEDLINE, Scopus, and Web of Science using predefined combinations of OSCC-, early-detection-, imaging-, cytology-, liquid-biopsy-, salivaomics-, artificial-intelligence-, and biosensor-related terms. English-language systematic reviews, meta-analyses, reviews of reviews, translational studies, and clinically relevant original articles were prioritized, with explicit attention to oral cavity-specific evidence and clearly identified extrapolation from broader head-and-neck or oropharyngeal cancer settings. Results: Conventional oral examination and histopathological assessment of biopsy specimens remain the diagnostic foundation. Adjunctive methods may support lesion triage, biopsy-site selection, risk stratification, and early monitoring, but cannot replace tissue diagnosis. Narrow-band imaging, optical coherence tomography, and molecular brush cytology appear particularly promising for specialist assessment and surveillance. Liquid biopsy and saliva-based biomarker platforms offer translational potential, particularly for repeatable monitoring, but clinical implementation is limited by methodological heterogeneity, pre-analytical variability, inconsistent thresholds, and insufficient external validation. Artificial intelligence and biosensor platforms remain promising but largely developmental. Conclusions: Progress in early OSCC diagnosis and monitoring will most likely depend on integrated diagnostic models combining clinical examination, adjunctive imaging, minimally invasive sampling, molecular biomarkers, and computational decision-support tools, validated in prospective multicenter studies. Full article

Background/Objectives: Identifying aggressive tumor biology within early-stage hepatocellular carcinoma (HCC) remains challenging. Scores based on liver function, systemic inflammation, and HCC biomarkers have been linked to overall survival prognosis; however, the combined ability of these scores to assess tumor-specific prognosis in early-stage disease is unclear. In this single-center, prospective study, biomarker profiling with AFP, AFP-L3, and DCP, along with modified albumin–bilirubin (mALBI), and neutrophil–lymphocyte ratio (NLR)/platelet–lymphocyte ratios (PLRs) were evaluated to determine their prognostic role in assessing clinical manifestations of aggressive biology by stratifying HCC progression risk. Methods: Indices and biomarkers were assessed at BCLC-A-stage HCC diagnosis and prior to liver-directed therapy (LDT). The primary prospective study endpoint was time-to-advanced-stage tumor progression (TTP). Results: The cohort included 232 patients diagnosed with early-stage HCC who underwent treatment with LDT. A multivariate model revealed that mALBI grade (p = 0.021), cumulative lesion size (p = 0.005), and elevations in HCC biomarkers (p < 0.001) were associated with TTP. Biomarker profile stratified TTP (p < 0.001) in which patients with complex profiles (3+) had 1-year progression risks of 69%. The biomarker system retained the ability to stratify TTP within small (≤3 cm) and large (>3 cm) cumulative tumor burden (p < 0.001, p = 0.005). While PLR was not prognostic for TTP, NLR disappeared from the multivariate model and mALBI stratified long-term progression risk (p = 0.003). In low-complex biomarker patients (0–1+), mALBI stratified progression risk (p = 0.001). Conclusions: Multi-positive biomarker profiling in early-stage HCC identifies a population with clinical manifestations of aggressive tumor biology at high risk of rapid post-treatment disease progression that may benefit from more aggressive treatment approaches. In patients with low-risk biomarker profiles (0–1+), mALBI can assess longer-term (>1-year) post-treatment disease progression risk, while scores based on systemic inflammation were not associated with tumor-restricted outcomes. Full article

Osteoporosis (OP) and osteoarthritis (OA) are two common degenerative musculoskeletal disorders associated with aging and are traditionally classified and managed as distinct disease entities. Emerging evidence suggests that OP and OA may share bidirectional associations and common biological mechanisms, and that under specific pathological conditions they may develop into a mutually reinforcing comorbid state. The comorbidity of osteoporosis and osteoarthritis (OP–OA) is not a simple superimposition of bone loss and cartilage degeneration; rather, it represents a disorder of the osteochondral unit centered on disruption of the subchondral bone microenvironment. Alterations in the structural strength, remodeling dynamics, vascular and neural status, and bone marrow lesions of subchondral bone collectively reshape the local microenvironment, thereby directly affecting mechanical signal transmission and cellular behavior within the joint. Focusing on the subchondral bone microenvironment as the central pathological nexus, this review systematically summarizes how mechanical imbalance, aberrant bone remodeling, inflammatory activation, metabolic dysregulation, and cellular senescence jointly remodel the local niche in OP–OA comorbidity. These microenvironmental changes further induce phenotypic remodeling and fate deviation of bone marrow mesenchymal stem cells, bone remodeling-related cells, osteoimmune cells, and chondrocytes. On this basis, we integrate the regulatory roles of developmental signaling, mechanotransduction pathways, and inflammatory–immune signaling networks, and propose that microenvironment-driven cell fate plasticity may serve as a key mechanistic hub promoting the initiation and progression of OP–OA comorbidity as well as the persistent destabilization of the osteochondral unit. This perspective may help overcome the limitations of current studies that address OP and OA separately, and may provide a theoretical framework for early identification and stratification, biomarker discovery, and combined precision-targeted interventions for this comorbid condition. Full article

The accurate segmentation of gastric cancer cells is important in pathology for diagnosing and detecting diseases early. However, current approaches still suffer from limitations such as expensive annotation, fuzzy lesion boundaries, and weak feature expression. In order to solve these problems, we present MSAF-Net, a novel U-Net framework optimized both architecturally and in terms of the loss function. In particular, we incorporate a Multi-scale Dilated Pooling Fusion Block into the encoder stage to achieve enhanced interaction of multi-paths and thus improve features’ diversity and boundary sensitivity. We also introduce a Dual-Channel Attention Block in place of traditional convolution block in the decoder stage to restore better details and reconstruct the fuzzy boundaries. Meanwhile, a Diagonal Mahalanobis Consistency Loss is incorporated into our framework to facilitate class compactness. Experiments performed on the SEED-Gastric Carcinoma Stage 1 dataset show that the designed algorithm can reach 0.776 in Dice score and 0.821 in Accuracy, which outperforms the baseline method U-Net. It is clear that these results have shown the effectiveness and robustness of our proposed approach. The introduced algorithm allows for more precise quantification of gastric cancer cell morphology. Full article

Background: Multimodal breast ultrasound, including B-mode imaging, color Doppler flow imaging, and elastography, provides complementary information for lesion characterization. However, effectively integrating heterogeneous modalities remains challenging due to inconsistent feature distributions, limited cross-modal interaction, computational cost in existing methods, and sensitivity to noise and missing data. Methods: We presented an efficient Cross-Modal Interaction and Dynamic Fusion Network (CIDFNet) for multimodal breast ultrasound analysis. The framework integrates a multi-scale feature enhancement module to improve modality-specific representations, a cross-modal interaction module to enable early-stage feature exchange across modalities, and a dynamic fusion strategy to adaptively combine modality information based on feature reliability estimation. In addition, an invertible neural network is incorporated to reconstruct missing modality features during training. Results: Experiments on an internal dataset of 248 patients with 1532 images show that CIDFNet obtains an AUC of 85.69%, accuracy of 75.51%, recall of 50.00%, F1-score of 62.50%, and precision of 83.33%, while requiring 49.51 M parameters and 79.79 G FLOPs, respectively. Under a simplified Gaussian noise perturbation setting, performance degradation is observed. Conclusions: CIDFNet presents a framework for multimodal breast ultrasound analysis that reflects a trade-off between performance and computational efficiency. Full article

Swine Inflammation and Necrosis Syndrome (SINS) is a widespread condition in pigs and has been proposed as an early-life animal-based measure (ABM) for assessing health and welfare. However, its prognostic value for later-life outcomes under commercial conditions remains poorly understood. This study investigated the prevalence, progression, and predictive relevance of SINS in two organic pig farms in the Netherlands. Clinical SINS signs were assessed in suckling and weaned piglets and related to hematological parameters at weaning (35 weaned piglets) as well as tail integrity at slaughter. SINS lesions were highly prevalent in suckling piglets (approximately 80%) but markedly decreased after weaning. Lesion prevalence and severity differed substantially between farms and showed clear age-dependent patterns, peaking between days 3 and 5 of life. Higher SINS scores in suckling piglets were associated with systemic hematological alterations at weaning, including increased monocyte proportions, reduced platelet counts, and altered red blood cell indices. Importantly, early-life SINS was significantly associated with later tail integrity. Pigs with higher SINS scores showed a lower probability of intact tails at slaughter and subsequently a higher prevalence of tail lesions. These findings suggest that SINS may have potential as an early-life indicator of later tail outcomes; however, this hypothesis requires validation in larger studies involving a greater number of farms and production systems. Full article

Background/Objectives: Inflammation plays a central role in the pathophysiology of ST-elevation myocardial infarction (STEMI) and may influence myocardial injury, ventricular dysfunction and clinical outcomes. Simple inflammatory biomarkers derived from routine laboratory tests have been proposed as potential prognostic indicators in patients undergoing primary percutaneous coronary intervention (PCI). Objective: This study aimed to evaluate the association between admission inflammatory biomarkers, echocardiographic markers of ventricular dysfunction and in-hospital mortality in patients with STEMI treated with primary PCI. Methods: We conducted a retrospective observational study including 600 consecutive patients admitted with STEMI and treated with primary PCI between January 2021 and August 2025. Inflammatory biomarkers measured at admission included C-reactive protein (CRP); neutrophil-to-lymphocyte ratio (NLR); platelet-to-lymphocyte ratio (PLR); systemic immune-inflammation index (SII) and C-reactive protein-to-lymphocyte ratio (CLR). Echocardiographic parameters and clinical outcomes were recorded. Multivariable logistic regression analysis was performed to identify independent predictors of in-hospital mortality. Results: In-hospital mortality occurred in 54 patients (9.0%). Patients with reduced left ventricular ejection fraction (LVEF ≤ 40%) had significantly higher CRP and CLR levels (p < 0.01). Inflammatory biomarkers were associated with markers of ventricular dysfunction but were not independent predictors of mortality. Age, LVEF < 40% and the number of residual coronary lesions independently predicted in-hospital death. Conclusions: In STEMI patients undergoing primary PCI, early mortality is mainly determined by age; ventricular dysfunction and residual coronary disease burden, while inflammatory biomarkers primarily reflect the severity of myocardial injury rather than independently predicting short-term mortality. Full article