Search Results (7,074)

Endoscopic submucosal dissection (ESD) has revolutionized the management of early upper gastrointestinal (GI) carcinomas. While technically demanding, it offers, in experienced hands, definitive local therapy for early GI neoplasia by allowing complete En bloc resection of mucosal and superficially invasive neoplasms, thus enabling precise histopathological risk stratification and organ preservation. Appropriate patient selection relies on meticulous endoscopic assessment using high-definition and image-enhanced endoscopy to define lesion boundaries and predict invasion depth. The principal indications include high-grade intraepithelial neoplasia and early carcinomas without endoscopic evidence of deep submucosal invasion or lymph node metastasis risk factors. Pathological analysis of the resection specimens includes histological typing and grading per WHO classification and precise assessment of invasion depth—in case of submucosal invasion measurement in micrometers—and evaluation of margin status and lymphovascular invasion. The presence of risk factors such as deep invasion in the submucosa, poor differentiation, or lymphovascular invasion may require additional surgery, guided by validated risk scores such as the eCura system. This narrative review summarizes current clinical and pathological practices for ESD in upper GI lesions. This includes the discussion of technical and biological challenges and the need of accurate assessment of risk factors for systemic metastatic spread and local recurrence as a limitation for this sophisticated but highly effective therapeutic method. Full article

Next-generation space observatories for high-energy gamma-ray astrophysics will increase scientific return using onboard machine learning (ML). This is now possible thanks to today’s low-power, radiation-tolerant processors and artificial intelligence accelerators. This paper provides an overview of current and future ML applications in gamma-ray space missions focused on high-energy transient phenomena. We discuss onboard ML use cases that will be implemented in the future, including real-time event detection and classification (e.g., gamma-ray bursts), and autonomous decision-making, such as rapid repointing to transient events or optimising instrument configuration based on the scientific target or environmental conditions. Full article

Background: Ovarian cancer is characterized by high mortality rates, primarily due to diagnosis at late stages. Current biomarkers, such as CA125, have demonstrated limited efficacy for early detection. While high-dimensional proteomics offers a more comprehensive view of systemic biology, the analysis of such data, where the number of features far exceeds the number of samples, presents a significant computational challenge. Methods: This study utilized a nested case–control cohort of longitudinal pre-diagnostic serum samples from the UK Collaborative Trial of Ovarian Cancer Screening (UKCTOCS) profiled for eight candidate ovarian cancer biomarkers (CA125, HE4, PEBP4, CHI3L1, FSTL1, AGR2, SLPI, DNAH17) and 92 additional cancer-associated proteins from the Olink Oncology II panel. We employed a Synolitic Graph Neural Network framework that transforms high-dimensional multi-protein data into sample-specific, interconnected graphs using a synolitic network approach. These graphs, which encode the relational patterns between proteins, were then used to train Graph Neural Network (GNN) models for classification. Performance of the network approach was evaluated together with conventional machine learning approaches via 5-fold cross-validation on samples collected within one year of diagnosis and a separate holdout set of samples collected one to two years prior to diagnosis. Results: In samples collected within one year of ovarian cancer diagnosis, conventional machine learning models—including XGBoost, random forests, and logistic regression—achieved the highest discriminative performance, with XGBoost reaching an ROC-AUC of 92%. Graph Convolutional Networks (GCNs) achieved moderate performance in this interval (ROC-AUC ~71%), with balanced sensitivity and specificity comparable to mid-performing conventional models. In the 1–2 year early-detection window, conventional model performance declined sharply (XGBoost ROC-AUC 46%), whereas the GCN maintained robust discriminative ability (ROC-AUC ~74%) with relatively balanced sensitivity and specificity. These findings indicate that while conventional approaches excel at detecting late pre-diagnostic signals, GNNs are more stable and effective at capturing subtle early molecular changes. Conclusions: The synolitic GNN framework demonstrates robust performance in early pre-diagnostic detection of ovarian cancer, maintaining accuracy where conventional methods decline. These results highlight the potential of network-informed machine learning to identify subtle proteomic patterns and pathway-level dysregulation prior to clinical diagnosis. This proof-of-concept study supports further development of GNN approaches for early ovarian cancer detection and warrants validation in larger, independent cohorts. Full article

Among cardiomyopathies, the hypertrophic phenotype is the most common, and hypertrophic cardiomyopathy (HCM) phenocopies represent a heterogeneous group of conditions. They are defined by a left ventricular wall thickness ≥15 mm in the absence of other causes such as loading conditions, ischemia, or valvular disease. Although they mimic similar clinical and morphological features, their etiologies are distinct and include genetic, metabolic, and infiltrative mechanisms. Therefore, accurate classification and differential diagnosis are crucial for effective management and treatment. Sarcomeric HCM is the most frequent form, accounting for up to 60% of cases. However, numerous non-sarcomeric phenocopies exist, including amyloidosis, Fabry disease, glycogen storage disorders, RASopathies, and mitochondrial diseases. Clinical and imaging findings are essential to distinguish these entities from sarcomeric HCM. Electrocardiography, echocardiography, advanced modalities such as cardiac magnetic resonance (CMR), and specific laboratory tests all play a central role in guiding diagnosis. Genetic testing provides key insights into mutations and inheritance patterns, further supporting definitive diagnosis. Correct identification of an HCM phenocopy carries important therapeutic implications, as disease-specific treatments can significantly improve prognosis. For example, targeted therapies exist for amyloidosis, Fabry disease, and certain metabolic or mitochondrial disorders, underlining the clinical relevance of an accurate diagnosis. This review aims to provide an overview of HCM phenocopies and assist clinicians in diagnostic reasoning. The first part addresses classification according to pathophysiological mechanisms, clinical features, and genetic background. The second part focuses on the stepwise approach to differential diagnosis, integrating clinical assessment, laboratory evaluation, ECG, echocardiography, and CMR findings. Full article

Background/Objectives: Bioinductive bovine collagen implants (BCI) have been introduced to enhance tendon biology and promote tissue regeneration in rotator cuff (RC) repairs. This study aimed to assess the clinical and radiological outcomes of arthroscopic posterosuperior rotator cuff (psRC) repair with BCI augmentation in full-thickness tears at increased risk of retear. Methods: This case series analyzed 30 patients with psRC tears who were classified as being at high risk of failure according to a predefined set of parameters, including patient history, radiological findings and intraoperative assessments, and the presence of psRC retears. All patients subsequently underwent arthroscopic psRC repair with BCI augmentation, compromising 21 primary and 9 secondary repairs. Clinical outcomes were assessed using Subjective Shoulder Value (SSV), American Shoulder and Elbow Surgeons (ASES) shoulder score, and Constant score at 6 and 12 months postoperatively. Tendon integrity was assessed using the Sugaya classification. Results: At 12 months, magnetic resonance imaging revealed complete tendon healing in 56.7%, partial healing in 16.7%, and insufficient healing in 26.7%. Significant improvements in SSV (45.3 to 83.5), ASES (40.6 to 77.8), and Constant score (36.6 to 71.7) were observed at 12 months postoperatively, with all outcome measures exceeding their respective minimally clinically important differences. Two patients (6.7%) developed secondary shoulder stiffness, and 1 patient (3.3%) required revision surgery for bicipital groove pain. Conclusions: Augmentation with a BCI in arthroscopic repair of high-risk psRC tears demonstrate promising short-term results. Patients achieve significant improvements in pain and shoulder function, accompanied by satisfactory tendon healing on MRI. Full article

The rise in thermal runaway events within electric vehicle (EV) battery systems requires anticipatory models to predict critical safety failures during operation. This investigation develops a multi-physics digital twin framework that links electrochemical, thermal, and structural domains to replicate the internal dynamics of lithium-ion packs in both normal and faulted modes. Coupled simulations distributed among MATLAB 2024a, Python 3.12-powered three-dimensional visualizers, and COMSOL 6.3-style multi-domain solvers supply refined spatial resolution of temperature, stress, and ion concentration profiles. While the digital twin architecture is designed to accommodate different battery chemistries and pack configurations, the numerical results reported in this study correspond specifically to a lithium NMC-based 4S3P cylindrical cell module. Quantitative benchmarks show that the digital twin identifies incipient thermal deviation with 97.4% classification accuracy (area under the curve, AUC = 0.98), anticipates failure onset within a temporal margin of ±6 s, and depicts spatial heat propagation through three-dimensional isothermal surface sweeps surpassing 120 °C. Mechanical models predict casing strain concentrations of 142 MPa, approaching polymer yield strength under stress load perturbations. A unified operator dashboard delivers diagnostic and prognostic feedback with feedback intervals under 1 s, state-of-health (SoH) variance quantified by a root-mean-square error of 0.027, and mission-critical alerts transmitting with a mean latency of 276.4 ms. Together, these results position digital twins as both diagnostic archives and predictive safety envelopes in the evolution of next-generation EV architectures. Full article

Modelling connectivity in utility networks is essential for operational management, maintenance planning, and resilience analysis. The CityGML Utility Network Application Domain Extension (UNADE) provides a detailed conceptual framework for representing utility networks; however, most existing implementations rely on relational databases, where connectivity must be reconstructed through joins rather than represented as explicit relationships. This creates challenges when managing densely connected network structures. This study introduces the UNADE–Labelled Property Graph (UNADE-LPG) model, a graph-based representation that maps the classes, relationships, and constraints defined in the UNADE Unified Modelling Language (UML) schema into nodes, edges, and properties. A conversion pipeline is developed to generate UNADE-LPG instances directly from CityGML UNADE datasets encoded in GML, enabling the population of graph databases while maintaining semantic alignment with the original schema. The approach is demonstrated through two case studies: a schematic network and a real-world water system from Frankston, Melbourne. Validation procedures, covering structural checks, topological continuity, classification behaviour, and descriptive graph statistics, confirm that the resulting graph preserves the semantic structure of the UNADE schema and accurately represents the physical connectivity of the network. An analytical path-finding query is also implemented to illustrate how the UNADE-LPG structure supports practical network-analysis tasks, such as identifying connected pipeline sequences. Overall, the findings show that the UNADE-LPG model provides a clear, standards-aligned, and operationally practical foundation for representing utility networks within graph environments, supporting future integration into digital-twin and network-analytics applications. Full article

This paper studies the symmetries of the extension to three points of the Dubins problem, the Three-Point Dubins Problem (3PDP), which consists of finding the shortest curvature-constrained $C^1$ path passing through three waypoints, which are the first and last oriented. In the literature, the optimal solution is selected by enumerating 18 possible candidates: the best is elected as the global solution of the instance of the 3PDP. To reduce the need of this enumeration, we exploit the symmetries of the problem to improve the solution strategy by using a Machine Learning (ML) framework. We show how to map an arbitrary configuration into a canonical domain and significantly reduce the parameter space, without a loss of generality. Then, we use this method to construct a compact yet comprehensive dataset of over 17 million valid cases. The reduction in the input dimensionality leads to a faster and more robust learning approach; we investigate both regression and classification neural networks, where the regression model estimates the optimal intermediate angle, and the classification model predicts the path type. The classification network achieved a top-1 accuracy of 97.5% and 100% accuracy within the top-5 predictions (instead of testing all 18 cases), whereas the regression model attained a mean angular error of about $2^{°}$. A detailed case study illustrates how the proposed method can complement existing analytic approaches by providing accurate initial guesses, thus accelerating iterative solvers. Our results demonstrate that ML-based methods can serve as efficient and reliable alternatives for solving the 3PDP, with direct implications for other motion planners in robotics and autonomous systems. Full article

Background: Climate change, urbanization, and global mobility increase the risk of emerging infectious diseases with pandemic potential. There is a need for rapid methods that can assess their long-term effects on human health. In silico approaches are particularly suited to study processes that may manifest years later, under the assumption that perturbed biomolecular interactions underlie these outcomes. Here we focus on viral oncogenicity—the ability of viruses to increase cancer risk—which accounts for about 15% of global cancer cases. Methods: We characterize viruses through multilayer representations of protein–protein interaction (PPI) networks reconstructed from the human interactome. Statistical analyses of topological features, combined with interpretable machine learning models, are used to distinguish oncogenic from non-oncogenic viruses and to identify proteins with potential central role in these processes. Results: Our analysis reveals clear statistical differences between the network properties of oncogenic and non-oncogenic viruses. Furthermore, the machine learning approach enables classification of virus–host interaction networks and identification of relevant subsets of proteins associated with oncogenesis. Functional enrichment analysis highlights mechanisms related to viral oncogenicity, including chromatin structure and other processes linked to cancer development. Conclusions: This framework enables virus classification and highlights mechanisms underlying viral oncogenicity, providing a foundation for investigating long-term health effects of emerging pathogens. Full article

Honey authenticity, including its botanical origin, is traditionally assessed by melissopalynology, a labour-intensive and expert-dependent method. This study reports the final validation of a deep learning model for pollen grain classification in honey, developed within the NUTRITECH.I-004A/22 project, by comparing its performance with that of an independent palynology expert. A dataset of 5194 pollen images was acquired from five unifloral honeys, rapeseed (Brassica napus), sunflower (Helianthus annuus), buckwheat (Fagopyrum esculentum), phacelia (Phacelia tanacetifolia) and linden (Tilia cordata), under a standardized microscopy protocol and manually annotated using an extended set of morphological descriptors (shape, size, apertures, exine ornamentation and wall thickness). The evaluation involved training and assessing a deep learning model based solely on the ResNet152 architecture with pretrained ImageNet weights. This model was enhanced by adding additional layers: a global average pooling layer, a dense hidden layer with ReLU activation, and a final softmax output layer for multi-class classification. Model performance was assessed using multiclass metrics and agreement with the expert, including Cohen’s kappa. The AI classifier achieved almost perfect agreement with the expert (κ ≈ 0.94), with the highest accuracy for pollen grains exhibiting spiny ornamentation and clearly thin or thick walls, and lower performance for reticulate exine and intermediate wall thickness. Misclassifications were associated with suboptimal image quality and intermediate confidence scores. Compared with traditional melissopalynological assessment (approx. 1–2 h of microscopic analysis per sample), the AI system reduced the effective classification time to less than 2 min per prepared sample under routine laboratory conditions, demonstrating a clear gain in analytical throughput. The results demonstrate that, under routine laboratory conditions, AI-based digital palynology can reliably support expert assessment, provided that imaging is standardized and prediction confidence is incorporated into decision rules for ambiguous cases. Full article

Background/Objectives: Some personalized alignment (PA) concepts have been described with symmetrical gaps in extension and flexion. However, laxity in native knees was significantly greater laterally than medially with respect to both extension and flexion. We hypothesized that a personalized alignment can restore the native knee alignment, keep a satisfying patellar tracking, and obtain physiological ligament balancing, that is, a symmetric gap in extension and an asymmetric gap in flexion. We aimed to assess: (1) the postoperative alignment of TKA and postoperative patellar tracking (primary outcome); (2) the ligament balancing at the end of the surgery; and (3) clinical outcomes and complication rates. Methods: In this single-center, retrospective case series, we evaluated 45 patients in a consecutive series who underwent robotic-assisted primary TKA using PA between January and September 2020 with a minimum follow-up of 5 years. Complication was defined as grade ≥3 according to the Clavien-Dindo classification. Data assessed were: TKA alignment and implant positioning on postoperative radiographs, patellar tracking on the merchant view, and ligament balancing in extension and flexion upon completion of surgery. Results: Mean follow-up was 62.1 ± 2.5 months. The postoperative mean HKA angle was 177.4° ± 2.2. The medial distal femoral angle was restored (91.1° ± 1.5 postoperatively versus 91.3° ± 2). A total of four TKAs had a patellar tilt superior to 5° (8.9%). No significant difference was found in the medial gap laxity—both in extension and in flexion—and the lateral gap laxity in extension. The lateral gap laxity in flexion was significantly higher than extension or medial gap laxity (+2.9 mm). One patient was readmitted for delayed wound healing. Average improvements in Knee Society knee and function scores were 55.86 and 51.84 points, respectively. Conclusions: This personalized alignment technique using anatomo-functional implant positioning allowed restoration of native knee alignment with a “safe zone” (3° varus/valgus) for the tibial implant, maintained satisfying patellar tracking, and restituted the asymmetrical gap laxity in flexion with a higher laxity in the lateral compartment. Being the longest system-specific study to date, the results are encouraging at 5 years with no major complications. However, longer follow-up will be required to confirm the use of this technique. Full article

Background/Objectives: Enthesitis is a hallmark feature across the spondylarthritis spectrum, including psoriatic arthritis (PsA). In recent years, advanced imaging techniques, particularly musculoskeletal ultrasound (MSUS), have demonstrated higher sensitivity than clinical examination in detecting enthesitis. This study aimed to evaluate the inter-observer agreement for the diagnosis of Achilles enthesitis in a cohort of PsA patients. A secondary objective was to explore specific ultrasound diagnostic criteria for identifying active, inflammatory enthesitis in this population. Methods: Adult patients with PsA, all fulfilling CASPAR classification criteria, were recruited and underwent both clinical and ultrasonographic assessment of the bilateral Achilles tendons. Each patient was scanned by 4 rheumatologists in a direct study, followed by a blinded evaluation of static images of the same patients. The examiners assessed the presence of enthesitis components according to the OMERACT criteria. In addition, the images were subsequently evaluated by 10 MSUS-experienced rheumatologists who were asked to classify the enthesitis as inflammatory by selecting one of the following responses: “yes”, “no,” or “possible”. Results: Ten PsA patients, with a median age of 60 and a median DAPSA score of 21, were included. Both direct and image-based inter-observer studies showed high agreement values for enthesophytes (κ > 0.6), erosions (κ > 0.5), and entheseal thickness (κ > 0.5). In both, low agreement was observed for hypoechogenicity (κ between 0.1 and 0.4). Erosions and power Doppler (PD) signal in erosions showed statistically significant differences between the “possible” and definite (“yes”) inflammatory enthesitis groups. A PD signal of grade 2 or 3 within the enthesis or erosions was observed exclusively in cases classified as definite (“yes”) inflammatory enthesitis. Similarly, a grade 3 PD signal in the bursa was found only in patients with definite inflammatory enthesitis. This study proposes a novel ultrasound scoring system for defining inflammatory enthesitis. The score demonstrated overall good diagnostic performance, with a sensitivity of 67% and a specificity of 100%. Conclusions: The relatively low inter-observer agreement regarding hypoechogenicity and the presence of PD highlights the need for targeted educational interventions to improve interpretation in MSUS. Erosions and PD signal within erosions appear to be significant discriminatory features for identifying inflammatory enthesitis. Full article

Against the backdrop of insufficient accuracy and adaptability of satellite precipitation products in complex terrain areas, this study focused on the Min River Basin (MRB) on the eastern edge of the Qinghai–Tibet Plateau. A two-step machine learning fusion framework was established, which integrates precipitation event identification and quantitative intensity estimation in a systematic manner. This framework incorporated 5 precipitation products (PERSIANN-CDR, CMORPH, GSMaP, IMERG, MSWEP), measured data, and environmental variables. The study compared the precipitation estimation performance of Random Forest (RF), Extreme Learning Machine (ELM), eXtreme Gradient Boosting (XGBoost), Bagging, and Double Machine Learning (DML) models, and analyzed the models’ performance under different precipitation intensities and altitudes, as well as their variable sensitivity. The results showed that: (1) DML models outperformed Single Machine Learning (SML) models and original precipitation products, with RF-Bagging being the optimal model. The daily-scale Correlation Coefficient (CC) of RF-Bagging was over 50% higher than that of original products, while the Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) were reduced by more than 40% and 35%, respectively. (2) For moderate-to-heavy precipitation, the RF-Bagging and RF-RF models maintain a stable Critical Success Index (CSI) of 0.7. In high-altitude regions, their Probability of Detection (POD) approaches 1, and the Heidke Skill Score (HSS) is 30–40% higher than that in mid-altitude areas, significantly outperforming other models and demonstrating strong adaptability to complex terrain. For light precipitation, while the POD values of these two models are comparable to those of other models, their False Alarm Rate (FAR) is reduced by 15–20%, effectively mitigating precipitation false alarms. (3) GSMaP, IMERG, and MSWEP were the core input variables for all models. RF and ELM models were more dependent on environmental variables, while XGBoost and Bagging models relied more on satellite data. This framework can provide technical references for precipitation estimation in complex terrain areas and contribute to watershed water resource management as well as flood prevention and mitigation. Full article

Operation and maintenance (O&M) events resulting from environmental factors (e.g., precipitation, temperature, seasonality, and unexpected weather conditions) are among the primary sources of operating costs and downtime in run-of-river small hydropower plants (SHPs). This paper presents a data-driven methodology for predicting such long events using machine learning models trained on historical power production, weather radar, and forecast data. Case studies on two Slovenian SHPs with different structural designs and levels of automation demonstrate how environmental features—such as day of year, rain duration, cumulative amount of rain, and rolling precipitation sums—can be used to forecast long events or shutdowns. The proposed approach integrates probabilistic classification outputs with threshold-consistency smoothing to reduce noise and stabilize predictions. Several algorithms were tested—including Logistic Regression, Support Vector Machine (SVM), Random Forest, Gradient Boosting, and k-Nearest Neighbors (k-NN)—across varying feature combinations for O&M model development, with cross-validation ensuring robust evaluation. The models achieved an F1-score of up to 0.58 in SHP1 (k-NN), showing strong seasonality dependence, and up to 0.68 in SHP2 (Gradient Boosting). For SHP1, the best model (k-NN) correctly detected 36 long events, while 15 were misclassified as no events and 38 false alarms were produced. For SHP2, the best model (Gradient Boosting) correctly detected 69 long events, misclassified 23 as no events, and produced 42 false alarms. The findings highlight that probabilistic machine learning-based forecasting can effectively support predictive O&M planning, particularly for manually operated or service-operated SHPs. Full article

This study develops an automatic dispatch system for quadrotor UAVs that integrates air-writing gesture recognition with a graphical user interface (GUI). The DJI RoboMaster quadrotor UAV (DJI, Shenzhen, China) was employed as the experimental platform, combined with an ESP32 microcontroller (Espressif Systems, Shanghai, China) and the RoboMaster SDK (version 3.0). On the Python (version 3.12.7) platform, a GUI was implemented using Tkinter (version 8.6), allowing users to input addresses or landmarks, which were then automatically converted into geographic coordinates and imported into Google Maps for route planning. The generated flight commands were transmitted to the UAV via a UDP socket, enabling remote autonomous flight. For gesture recognition, a Raspberry Pi integrated with the MediaPipe Hands module was used to capture 16 types of air-written flight commands in real time through a camera. The training samples were categorized into one-dimensional coordinates and two-dimensional images. In the one-dimensional case, X/Y axis coordinates were concatenated after data augmentation, interpolation, and normalization. In the two-dimensional case, three types of images were generated, namely font trajectory plots (T-plots), coordinate-axis plots (XY-plots), and composite plots combining the two (XYT-plots). To evaluate classification performance, several machine learning and deep learning architectures were employed, including a multi-layer perceptron (MLP), support vector machine (SVM), one-dimensional convolutional neural network (1D-CNN), and two-dimensional convolutional neural network (2D-CNN). The results demonstrated effective recognition accuracy across different models and sample formats, verifying the feasibility of the proposed air-writing trajectory framework for non-contact gesture-based UAV control. Furthermore, by combining gesture recognition with a GUI-based map planning interface, the system enhances the intuitiveness and convenience of UAV operation. Future extensions, such as incorporating aerial image object recognition, could extend the framework’s applications to scenarios including forest disaster management, vehicle license plate recognition, and air pollution monitoring. Full article