Search Results (2,498)

Evaluation of pituitary neuroendocrine tumors remains complex depending on the exact growth pattern, involvement of critical neurovascular structures, pituitary function and endocrinological activity of the tumor. A predominant growth into the sphenoid sinus and clivus poses specific challenges. We reviewed 557 surgeries for pituitary neuroendocrine tumors in an endonasal endoscopic technique performed between 1 January 2015 and 31 August 2025 to identify 13 cases (2.3%). Clinical, radiological and surgical data were selected by chart review. Thirteen cases aged from 31 to 68 years with almost exclusively non-functioning or clinically silent tumors (92%) were identified. Clival infiltration was restricted to the dorsum sellae in 2/13 (15%), spread to the floor of the sphenoid in 6/13 (46%) and extended inferior to the sphenoid in 5/13 (38%) cases with a high rate of cavernous sinus (62%) and sphenoid sinus infiltration (69%). Complete resection was achieved in 31%, and the residual tumor was clival/sphenoidal in 5/13 cases or within the cavernous sinus in 6/13 cases. The diaphragma sellae was reported to be intact in 92% of cases, and postoperative transient arginine vasopressin deficiency did not occur. Pituitary neuroendocrine tumors predominantly growing below the sella and infiltrating the clivus and sphenoid present specific challenges with a high rate of preoperative pituitary insufficiency, frequent cavernous sinus infiltration and postoperative tumor residuals in the cavernous sinus, sphenoid bone and clivus which are sometimes difficult to delineate. The surgical approach must be tailored specifically to treat the clival infiltration zone to reduce the risk of recurrence. Full article

Recent advances in automated learning techniques have enhanced the analysis of biomedical signals for detecting sleep stages and related health abnormalities. However, many existing models face challenges with imbalanced datasets and the dynamic nature of evolving sleep states. In this study, we present a robust algorithm for classifying sleep states using electroencephalogram (EEG) data collected from 33 healthy participants. We extracted dynamic, brain-inspired features, such as microstates and Lempel–Ziv complexity, which replicate intrinsic neural processing patterns and reflect temporal changes in brain activity during sleep. An optimal feature set was identified based on significant spectral ranges and classification performance. The classifier was developed using a convolutional neural network (CNN) combined with gated recurrent units (GRUs) within a reinforcement learning framework, which models adaptive decision-making processes similar to those in biological neural systems. Our proposed biomimetic framework illustrates that a multivariate feature set provides strong discriminative power for sleep state classification. Benchmark comparisons with established approaches revealed a classification accuracy of 98% using the optimized feature set, with the framework utilizing fewer EEG channels and reducing processing time, underscoring its potential for real-time deployment. These findings indicate that applying biomimetic principles in feature extraction and model design can improve automated sleep monitoring and facilitate the development of novel therapeutic and diagnostic tools for sleep-related disorders. Full article

Human microbiota, a complex consortium of microorganisms co-evolved with the host, profoundly influences tissue development, immune regulation, and disease progression. Growing evidence shows that microbial metabolites and signaling molecules modulate key stem cell pathways—such as Hedgehog, Wnt/β-catenin, and Notch—thereby reprogramming stem cell fate toward tumor-suppressive or tumor-promoting outcomes. Specific taxa within oral, intestinal, and urogenital niches have been linked to cancer initiation, therapy resistance, and recurrence. In parallel, clinical studies reveal that microbiota composition affects infection dynamics: bacterial isolates from symptomatic urinary tract infections inhibit commensal growth more strongly than the reverse, with Gram-positive and Gram-negative strains displaying distinct interaction profiles. Collectively, these findings highlight microbiota’s dual role in regulating cellular plasticity and pathogenicity. Elucidating host–microbe and microbe–microbe mechanisms may guide microbiota-targeted interventions to improve cancer and infectious disease management. Full article

Hyperspectral imaging (HSI) captures the same scene across multiple spectral bands, providing richer spectral characteristics of materials than conventional RGB images. The spectral reconstruction task seeks to map RGB images into hyperspectral images, enabling high-quality HSI data acquisition without additional hardware investment. Traditional methods based on linear models or sparse representations struggle to effectively model the nonlinear characteristics of hyperspectral data. Although deep learning approaches have made significant progress, issues such as detail loss and insufficient modeling of spatial–spectral relationships persist. To address these challenges, this paper proposes the Cascaded Multi-Attention Feature Recurrent Enhancement Network (CMFREN). This method achieves targeted breakthroughs over existing approaches through a cascaded architecture of feature purification, spectral balancing and progressive enhancement. This network comprises two core modules: (1) the Hierarchical Residual Attention (HRA) module, which suppresses artifacts in illumination transition regions through residual connections and multi-scale contextual feature fusion, and (2) the Cascaded Multi-Attention (CMA) module, which incorporates a Spatial–Spectral Balanced Feature Extraction (SSBFE) module and a Spectral Enhancement Module (SEM). The SSBFE combines Multi-Scale Residual Feature Enhancement (MSRFE) with Spectral-wise Multi-head Self-Attention (S-MSA) to achieve dynamic optimization of spatial–spectral features, while the SEM synergistically utilizes attention and convolution to progressively enhance spectral details and mitigate spectral aliasing in low-resolution scenes. Experiments across multiple public datasets demonstrate that CMFREN achieves state-of-the-art (SOTA) performance on metrics including RMSE, PSNR, SAM, and MRAE, validating its superiority under complex illumination conditions and detail-degraded scenarios. Full article

This work investigates the deployment of Deep Learning (DL) models for network intrusion detection on resource-constrained IoT devices, using the public CICIoT2023 dataset. In particular, we consider the following DL models: Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), Recurrent Neural Network (RNN), Convolutional Neural Network (CNN), Temporal Convolutional Network (TCN), Multi-Layer Perceptron (MLP). Bayesian optimization is employed to fine-tune the models’ hyperparameters and ensure reliable performance evaluation across both binary (2-class) and multi-class (8-class, 34-class) intrusion detection. Then, the computational complexity of each DL model is analyzed—in terms of the number of Multiply–ACCumulate operations (MACCs), RAM usage, and inference time—through the STMicroelectronics Cube.AI Analyzer tool, with models being deployed on an STM32H7S78-DK board. To assess the practical deployability of the considered DL models, a trade-off score (balancing classification accuracy and computational efficiency) is introduced: according to this score, our experimental results indicate that MLP and TCN outperform the other models. Furthermore, Post-Training Quantization (PTQ) to 8-bit integer precision is applied, allowing the model size to be reduced by more than $90\%$ with negligible performance degradation. This demonstrates the effectiveness of quantization in optimizing DL models for real-world deployment on resource-constrained IoT devices. Full article

Background: Ameloblastomas account for roughly 1% of all jaw tumours and cysts, typically manifesting as slow-growing, painless swellings that expand both buccal and lingual cortical plates and may infiltrate adjacent soft tissue, often leading to a delayed diagnosis. These benign tumours, characterized by local invasiveness, originate from epithelial tissues and may develop from dental lamina cell rests, the enamel apparatus, the epithelial lining of odontogenic cysts, or basal epithelial cells of the oral mucosa. Methods: This paper aims to describe the comprehensive and interdisciplinary management of an extensive ameloblastoma in a 16-year-old patient, emphasizing the diagnostic challenges, surgical resection, reconstructive procedures, and subsequent oral rehabilitation. Results: At the eleven-year follow-up, clinical and radiographic examinations showed no signs of tumour recurrence. The patient presented no symptoms, indicating neither pain nor functional impairment. The prosthetic rehabilitation utilizing implant-supported fixed restorations was successfully completed, resulting in satisfactory masticatory function and aesthetics. This case adds to the existing evidence on the management of extensive ameloblastomas by demonstrating successful long-term outcomes following interdisciplinary surgical reconstruction and rehabilitation. Conclusions: The presented case highlights the complexity of restoring the lost tissues and functions, as well as the long-term clinical, functional, and aesthetic outcomes over an eleven-years follow-up period. Full article

Plasma cell-free RNA (cfRNA) metagenomics is increasingly explored for blood-based pathogen detection, but the structure of the shared background “blood microbiome”, the reproducibility of reported signals, and the practical limits of this approach remain unclear. We performed a critical re-analysis and benchmarking (“stress test”) of host-filtered blood RNA sequencing data from two cohorts: a bacteriologically confirmed tuberculosis (TB) cohort (n = 51) previously used only to derive host cfRNA signatures, and a coronary artery disease (CAD) cohort (n = 16) previously reported to show a CAD-shifted “blood microbiome” enriched for periodontal taxa. Both datasets were processed with a unified pipeline combining stringent human read removal and taxonomic profiling using the latest versions of specialized tools Kraken2 and MetaPhlAn4. Across both cohorts, only a minority of non-host reads were classifiable; under strict host filtering, classified non-host reads comprised 7.3% (5.0–12.0%) in CAD and 21.8% (5.4–31.5%) in TB, still representing only a small fraction of total cfRNA. Classified non-host communities were dominated by recurrent, low-abundance taxa from skin, oral, and environmental lineages, forming a largely shared, low-complexity background in both TB and CAD. Background-derived bacterial signatures showed only modest separation between disease and control groups, with wide intra-group variability. Mycobacterium tuberculosis-assigned reads were detectable in many TB-positive samples but accounted for ≤0.001% of total cfRNA and occurred at similar orders of magnitude in a subset of TB-negative samples, precluding robust discrimination. Phylogeny-aware visualization confirmed that visually “enriched” taxa in TB-positive plasma arose mainly from background-associated clades rather than a distinct pathogen-specific cluster. Collectively, these findings provide a quantitative benchmark of the background-dominated regime and practical limits of plasma cfRNA metagenomics for pathogen detection, highlighting that practical performance is constrained more by a shared, low-complexity background and sparse pathogen-derived fragments than by large disease-specific shifts, underscoring the need for transparent host filtering, explicit background modeling, and integration with targeted or orthogonal assays. Full article

Early and reliable detection of bearing faults is essential for ensuring the safe and efficient operation of rotating electrical machines, especially under varying loads and non-stationary operating conditions. However, traditional diagnostic approaches struggle to maintain accuracy when signals are noisy, high-dimensional, or affected by multiple fault patterns. To address these issues, this work presents RNN-XBoostNet, a lightweight hybrid framework that combines the temporal-feature extraction capability of Recurrent Neural Networks (RNNs) with the robust classification strength of XGBoost. A new feature-selection strategy, CoLaR-FS (integrating correlation analysis, Lasso regularization, and recursive feature elimination), is introduced to reduce redundancy and retain only the most discriminative fault features. The proposed framework is evaluated using the widely known CWRU dataset and a newly developed induction-machine dataset containing ten fault categories, including six newly introduced real-world conditions. Experimental results show significant performance improvements: accuracy increased from 87.01% to 99.35% on the CWRU dataset and from 79.98% to 99.57% on the laboratory dataset. The combination of high accuracy, reduced complexity, and strong generalization demonstrates that RNN-XBoostNet, supported by CoLaR-FS, is a practical and effective solution for modern condition-based monitoring systems. Full article

We describe a patient with recurrent, brief episodes of chest discomfort caused by a highly mobile papillary fibroelastoma originating from the aortic wall and intermittently encroaching on the right coronary artery ostium. Initial 2D and 3D transthoracic and 2D transesophageal echocardiography identified a highly mobile mass in the ascending aorta above the aortic valve; the exact site of attachment and its relationship to the coronary ostia could not be clearly defined. Three-dimensional transesophageal echocardiography enabled precise anatomical reconstruction of the lesion and surrounding structures, clearly demonstrating its pedicle and proximity to the right coronary ostium. This imaging modality clarified the pathophysiological mechanism of symptoms and facilitated optimal surgical planning without the need for additional complex imaging techniques. Full article

In the context of carbon neutrality, the large-scale integration of renewable energy sources has led to frequent load changes in coal-fired boilers. These fluctuations cause key operational parameters to deviate significantly from their design values, undermining combustion stability and reducing operational efficiency. To address this issue, we introduce a novel predictive control method to enhance the control precision of key parameters under complex variable-load conditions, which integrates a coupled predictive model and real-time optimization. The predictive model is based on a coupled Transformer-gated recurrent unit (GRU) architecture, which demonstrates strong adaptability to load fluctuations and achieves high prediction accuracy, with a mean absolute error of 0.095% and a coefficient of determination of 0.966 for oxygen content (OC); 0.0163 kPa and 0.987 for bed pressure (BP); and 0.300 °C and 0.927 for main steam temperature (MST). These results represent substantial improvements over lone implementations of GRU, LSTM, and Transformer models. Based on these multi-step predictions, a WOA-based real-time optimization strategy determines coordinated adjustments of secondary fan frequency, slag discharger frequency, and desuperheating water valves before deviations occur. Field validation on a 300 t/h boiler over a representative 24 h load cycle shows that the method reduces fluctuations in OC, BP, and MST by 62.07%, 50.95%, and 40.43%, respectively, relative to the original control method. By suppressing parameter variability and maintaining key parameters near operational targets, the method enhances boiler thermal efficiency and steam quality. Based on the performance gain measured during the typical operating day, the corresponding annual gain is estimated at ~1.77%, with an associated CO₂ reduction exceeding 6846 t. Full article

One major factor influencing the development of eco-friendly policies and the implementation of climate change mitigation strategies is the accurate projection of CO₂ emissions. Traditional statistical models face significant limitations in capturing complex nonlinear interactions within high-dimensional emissions data. Advanced deep learning architectures offer new opportunities to overcome these computational challenges due to their strong pattern-recognition capabilities. This paper evaluates four distinct deep learning architectures for CO₂ emissions forecasting: Long Short-Term Memory (LSTM) networks, Gated Recurrent Units (GRUs), Hybrid Convolutional–LSTM (CNN–LSTM) systems, and Dense Neural Networks (DNNs). A comprehensive comparison is conducted using consistent training protocols, hyperparameters, and performance metrics across five prediction horizons (1, 3, 6, 12, and 24 steps ahead) to reveal architecture-specific degradation patterns. Furthermore, analyzing emissions by category provides insight into the suitability of each architecture for varying levels of pattern complexity. LSTM-based models demonstrate particular strength in modeling long-term temporal dependencies, making them well-suited for integration into long-range environmental policy planning frameworks. Overall, this study provides empirical evidence supporting the use of neural networks in climate modeling and proposes criteria for selecting optimal architectures based on forecasting horizon and computational constraints. Full article

Cancer remains one of the main global public health challenges, with rising incidence and mortality rates demanding more effective diagnostic and therapeutic approaches. Recent advances in artificial intelligence (AI) have positioned it as a transformative force in oncology, offering the ability to process vast and complex datasets that extend beyond human analytic capabilities. By integrating radiological, histopathological, genomic, and clinical data, AI enables more precise tumor characterization, including refined molecular classification, thereby improving risk stratification and facilitating individualized therapeutic decisions. In diagnostics, AI-driven image analysis platforms have demonstrated excellent performance, particularly in radiology and pathology. Prognostic algorithms are increasingly applied to predict survival, recurrence, and treatment response, while reinforcement learning models are being explored for dynamic radiotherapy and optimization of complex treatment regimens. Beyond direct patient care, AI is accelerating drug discovery and clinical trial design, reducing costs and timelines associated with translating novel therapies into clinical practice. Clinical decision support systems are gradually being integrated into practice, assisting physicians in managing the growing complexity of cancer care. Despite this progress, challenges such as data quality, interoperability, algorithmic bias, and the opacity of complex models limit widespread integration. Additionally, ethical and regulatory hurdles must be addressed to ensure that AI applications are safe, equitable, and clinically effective. Nevertheless, the trajectory of current research suggests that AI will play an increasingly important role in the evolution of precision oncology, complementing human expertise and improving patient outcomes. Full article

Central nervous system (CNS) tumors consist of a diverse set of malignancies that remain clinically challenging due to their biological complexity, high morbidity, and limited responsiveness to current therapies. A growing body of genomic evidence has revealed that dysregulation of the mitogen-activated protein kinase (MAPK) signaling pathway is a recurrent and unifying characteristic across many pediatric and adult CNS tumor entities. Alterations affecting upstream receptor tyrosine kinases (RTKs), RAS GTPases, RAF kinases, and other associated regulators contribute to MAPK signaling pathway hyperactivation, shaping tumor behavior, therapy response and resistance. These aberrations ranging from hotspot mutations such as BRAF V600E and oncogenic fusions like BRAF–KIAA1549 are particularly enriched in gliomas and glioneuronal tumors, highlighting MAPK signaling as a key oncogenic driver. The expanding availability of molecularly targeted compounds, including selective inhibitors of RAF, MEK and ERK, has begun to transform treatment approaches for specific molecular subtypes. However, the clinical benefit of MAPK-directed therapies is frequently limited by restricted blood–brain barrier (BBB) penetration, intratumoral heterogeneity, parallel pathway reactivation, and an immunosuppressive tumor microenvironment (TME). In this review, we synthesize current knowledge on MAPK pathway alterations in CNS tumors and evaluate the therapeutic landscape of MAPK inhibition, with emphasis on approved agents, emerging compounds, combination strategies, and novel drug-delivery technologies. We also discuss mechanisms that undermine treatment efficacy and highlight future directions aimed at integrating MAPK-targeted therapy into precision-based management of brain tumors. Full article

Background: The rising incidence of cutaneous non-melanoma skin cancers underscores the need for individualized reconstruction, particularly for cheek defects that pose distinctive anatomic and functional challenges. This study aimed to analyze reconstructive patterns for cheek skin lesions and to develop a simple, site- and size-based algorithm for small- to medium-sized defects. Methods: We retrospectively reviewed 129 consecutive patients treated between 2022 and 2025 for primary basal cell carcinoma, squamous cell carcinoma, or benign cheek skin tumors. After excision, defects were reconstructed with primary closure, local flaps, or skin grafts. Associations between the largest clinically measured lesion diameter (used as a proxy for the post-excision defect size), anatomical subsite, histopathology, and reconstructive technique were evaluated using ANOVA or Kruskal–Wallis tests, chi-square tests, and Spearman’s correlation. Results: The mean lesion diameter was 19.75 ± 12.93 mm. Reconstruction was performed using local flaps in 62 patients (48.06%), primary closure in 53 (41.09%), and skin grafts in 14 (10.85%). Larger defects were more frequently managed with grafts or flaps (F(2,110) = 4.84, p = 0.010), and lesion size correlated with reconstructive complexity (Spearman’s ρ = 0.229, p = 0.015). Lesion location was also significantly associated with the reconstruction method (χ²(10) = 48.29, p < 0.001; Cramér’s V = 0.44). Margin-negative (R0) excision was achieved in 95.35% of cases, with a low recurrence rate (3.91%) and complication rate (1.56%). Conclusions: Lesion size and anatomical location are key determinants of reconstructive strategy for cheek skin defects. In this cohort, lesions ≤ 20 mm were predominantly managed with primary closure, whereas lesions > 20 mm more frequently required flap reconstruction or skin grafting. This size-based split is cohort-derived and should be interpreted as a pragmatic framework that requires external validation. Full article

Seasonal influenza continues to pose a substantial and recurrent public health challenge in Queensland, driven by annual variability in transmission and uncertainty in climatic, demographic, and behavioural determinants. Predictive modelling is constrained by data limitations and parameter uncertainty. In response, this study developed a Bayesian network (BN) model to estimate the probability of influenza epidemics in Queensland, Australia. The model integrated diverse inputs, including international and local influenza surveillance data, demographic health statistics, and expert and stakeholder insights to capture the complex multifactorial causal relationships underlying epidemic risk. Scenario-based simulations revealed that Southeast Asian viral origin, severe global influenza seasons, peak season timing, increasing international travel, absence of control measures, and low immunisation rates substantially elevate the likelihood of influenza epidemics. Southeast Queensland was identified as particularly vulnerable under high-risk conditions. Model evaluation demonstrated good discriminative performance (AUC = 0.6974, accuracy = 70%) with appropriate uncertainty quantification through credible intervals and sensitivity analysis. Its modular design and capacity for integrating various data sources make it a practical decision-making support tool for public health preparedness and responding to evolving climatic and epidemiological conditions. Full article