Search Results (5,526)

Accurate brain tumor classification from MRI is often constrained by limited labeled data. We systematically compare conventional machine learning, deep learning, and few-shot learning (FSL) for four classes (glioma, meningioma, pituitary, no tumor) using a standardized pipeline. Models are trained on the Kaggle Brain Tumor MRI Dataset and evaluated across dataset regimes (100%→10%). We further test generalization on BraTS and quantify robustness to resolution changes, acquisition noise, and modality shift (T1→FLAIR). To support clinical trust, we add visual explanations (Grad-CAM/saliency) and report per-class results (confusion matrices). A fairness-aligned protocol (shared splits, optimizer, early stopping) and a complexity analysis (parameters/FLOPs) enable balanced comparison. With full data, Convolutional Neural Networks (CNNs)/Residual Networks (ResNets) perform strongly but degrade with 10% data; Model-Agnostic Meta-Learning (MAML) retains competitive performance (AUC-ROC ≥ 0.9595 at 10%). Under cross-dataset validation (BraTS), FSL—particularly MAML—shows smaller performance drops than CNN/ResNet. Variability tests reveal FSL’s relative robustness to down-resolution and noise, although modality shift remains challenging for all models. Interpretability maps confirm correct activations on tumor regions in true positives and explain systematic errors (e.g., “no tumor”→pituitary). Conclusion: FSL provides accurate, data-efficient, and comparatively robust tumor classification under distribution shift. The added per-class analysis, interpretability, and complexity metrics strengthen clinical relevance and transparency. Full article

Cordyceps militaris (C. militaris) is a medicinal fungus renowned for its diverse therapeutic properties, largely attributed to bioactive compounds such as cordycepin, polysaccharides, adenosine, D-mannitol, carotenoids, and ergosterol. However, the production and composition of these metabolites are highly influenced by cultivation conditions, highlighting the need for systematic optimization strategies. This review synthesizes current findings on how nutritional factors—including carbon and nitrogen sources, their ratios, and trace elements—and environmental parameters such as oxygen availability, pH, temperature, and light regulate C. militaris metabolite biosynthesis. The impacts of solid-state fermentation (using grains, insects, and agro-industrial residues) and liquid state fermentation (submerged and surface cultures) are compared, with attention to their roles in mycelial growth, fruiting body formation, and secondary metabolite production. Special emphasis is placed on mixed grain–insect substrates and light regulation, which have emerged as promising methods to enhance cordycepin accumulation. Beyond summarizing advances, this review also identifies key knowledge gaps that must be addressed: (i) the incomplete understanding of metabolite regulatory networks, (ii) the absence of standardized cultivation protocols, and (iii) unresolved challenges in scale-up, including oxygen transfer, foam control, and downstream processing. We propose that future research should integrate multi-omics approaches with bioprocess engineering to overcome these limitations. Collectively, this review highlights both current progress and remaining challenges, providing a roadmap for advancing the sustainable, scalable, and application-driven production of bioactive compounds from C. militaris. Full article

Penicillin G (PEN) is a widely used antibiotic for treating microbial infections. However, its extensive use in veterinary medicine can lead to accumulation in animal-derived products, particularly milk and meat. This highlights the urgent need for rapid and sensitive antibiotic detection methods. In this study, we employed DNA aptamers for the detection of PEN and for the analysis of aptamer specificity using a combined approach based on quartz crystal microbalance with dissipation monitoring (QCM-D) and localized surface plasmon resonance (LSPR). QCM-D measures changes in resonant frequency, Δf, and dissipation, ΔD, while LSPR monitors wavelength shifts in the extinction spectra corresponding to changes at the surface of gold nanoparticles (AuNPs). Thiolated aptamers were chemisorbed onto the surface of AuNPs with a diameter of 80 nm. In the presence of PEN, a redshift in the extinction spectra and a decrease in resonant frequency were observed, accompanied by an increase in dissipation due to surface viscosity effects. Significant changes in both acoustic and LSPR signals were observed at PEN concentrations as low as 1 nM. The limits of detection (LOD) for PEN, determined by QCM-D (3.0 nM, or 1.05 ng/mL)) and LSPR (3.1 nM, or 1.09 ng/mL), were similar and both were lower than the maximum residue limit (MRL) for PEN established by the EU (4 ng/mL). Full article

Virus-associated hepatocellular carcinoma (HCC) remains a major global health burden despite effective antiviral therapies. Chronic infection with hepatitis B (HBV), hepatitis C (HCV), and hepatitis D (HDV) promotes malignant transformation through overlapping pathways of fibrosis, immune dysregulation, and microenvironmental remodeling. YKL-40, a glycoprotein secreted by hepatic stellate cells, hepatocytes under stress, macrophages, and endothelial cells, has emerged as a marker that reflects stromal activation rather than direct hepatocyte injury. Its expression is reinforced by profibrotic and angiogenic circuits, and circulating concentrations correlate with advanced fibrosis, residual risk after viral suppression, and oncologic outcomes. This review synthesizes current evidence on YKL-40 across HBV, HCV, and HDV cohorts, with emphasis on its role in bridging molecular mechanisms to clinical applications. We examine its utility in non-invasive fibrosis assessment, longitudinal monitoring after antiviral therapy, and prognostic modeling in HCC. Particular attention is given to its potential in the liver transplant pathway, where YKL-40 may refine eligibility beyond morphology, inform bridging therapy response, and predict post-transplant recurrence or graft fibrosis. Remaining challenges include its lack of disease specificity, assay variability, and limited multicenter validation. Future integration of YKL-40 into multimarker, algorithm-based frameworks could enable risk-adaptive strategies that align surveillance and transplant decisions with the evolving biology of virus-associated liver disease. Full article

This study addresses the absence of maximum residue limits (MRLs) for tebuconazole and trifloxystrobin in edible rose petals in China by systematically evaluating the residue behavior and dietary exposure risks of these fungicides. An analytical method based on QuEChERS sample preparation coupled with UPLC–MS/MS was developed for the simultaneous determination of tebuconazole, trifloxystrobin, and its metabolite CGA321113 in fresh and dried rose petals. Field trials under the highest application conditions (184 g a.i./hm², applied twice) were conducted to investigate residue dissipation dynamics, storage stability, processing concentration effects, and transfer behavior during brewing. Results indicated that the target compounds remained stable in rose petals for 12 months at –20 °C ± 2 °C. The drying process significantly concentrated residues due to the hydrophobic nature of the compounds, with enrichment factors ranging from 3.0 to 3.9. Brewing tests further confirmed low transfer rates of tebuconazole, trifloxystrobin, and CGA321113 into the infusion, consistent with their low water solubility and high log K_ow values. Residue dissipation followed first-order kinetics, with half-lives of 1.9–2.9 days for tebuconazole and 1.2–2.7 days for trifloxystrobin. Dietary risk assessment showed an acceptable risk for trifloxystrobin (RQ = 22.7%) but a high risk for tebuconazole (RQ = 175.1%). It is recommended to set the MRL for both tebuconazole and trifloxystrobin in edible roses at 15.0 mg/kg. This standard ensures consumer safety while accommodating agricultural needs and aligns with international regulations. For the high-risk pesticide tebuconazole, measures such as optimizing application strategies and promoting integrated management should be implemented to mitigate residue risks. Full article

Extracellular signal-regulated kinase 1/2 (ERK1/2) inhibitors show therapeutic potential in triple-negative breast cancer (TNBC), but resistance through compensatory signaling limits their efficacy. We previously identified the secretory carrier membrane protein 3 (SCAMP3) as a regulator of TNBC progression and ERK1/2 activation. Here, we investigated the role of SCAMP3 in ERK1/2 signaling and therapeutic response using TMT-based LC-MS/MS phosphoproteomics of wild-type (WT) and SCAMP3 knockout (SC3KO) SUM-149 cells under basal conditions, after epidermal growth factor (EGF) stimulation, and during ERK1/2 inhibition with MK-8353. A total of 4408 phosphosites were quantified, with 1093 significantly changed. SC3KO abolished residual ERK activity under MK-8353 and affected the compensatory activation of oncogenic pathways observed in WT cells. SC3KO reduced the phosphorylation of ERK feedback regulators RAF proto-oncogene serine/threonine-protein kinase Raf-1 (S43) and the dual-specificity mitogen-activated protein kinase kinase 2 (MEK2) (T394), affected other ERK targets, including nucleoporins, transcription factors, and metabolic enzymes triosephosphate isomerase (TPI1) (S21) and ATP-citrate lyase (ACLY) (S455). SCAMP3 loss also impaired the mammalian target of rapamycin complex I (mTORC1) signaling and disrupted autophagic flux, evidenced by elevated sequestosome-1 (SQSTM1/p62) and microtubule-associated protein light chain 3 (LC3B-II) with reduced levels of the autophagosome lysosome maturation marker, Rab7A. Beyond ERK substrates, SC3KO affected phosphorylation events mediated by other kinases. These findings position SCAMP3 as a central coordinator of ERK signaling and autophagy. Our results support SCAMP3 as a potential therapeutic target to enhance ERK1/2 inhibitor clinical efficacy and overcome adaptive resistance mechanisms in TNBC. Full article

Global food systems face growing pressure from population expansion and climate change, making the identification of resilient crops a priority. The nopal cactus (Opuntia spp.) stands out for its capacity to thrive in arid environments and for its cultural and economic importance in Mexico. This study analyzes worldwide research trends and evaluates evidence from Mexico to identify opportunities and strategies for closing production cycles through residue valorization. Scientific output over the past decade shows steady growth and a thematic transition from basic agronomic and compositional studies toward sustainability, bioactive compounds, and circular economy approaches. In the Mexican context, applied studies demonstrate that Opuntia spp. cladodes residues can be transformed into composts with C/N ratios between 12 and 26, improving soil organic matter and nutrient availability. Biofertilizers produced through anaerobic fermentation enhanced phosphorus solubility in alkaline soils, while direct residue incorporation increased carrot and tomato yields up to threefold. Farmers recognize these practices as low-cost and compatible with local systems. Nevertheless, the lack of standardized protocols and scalable models limits widespread adoption. Strengthening research collaboration, policy incentives, and technology transfer could position Mexico as a leader in sustainable Opuntia value chains and advance circular economy practices in smallholder farming systems. Full article

Animal farming produces large volumes of underutilised by-products, such as poultry feathers (PF), often discarded in landfills or incinerated, causing environmental concerns. Transforming such residues into valuable energy carriers aligns with sustainable waste-to-energy (WtE) management. Pyrolysis represents a versatile thermochemical pathway for converting organic wastes into gaseous, liquid, and solid fuels. This study investigates slow pyrolysis of PF, lignite (LG), and their blends at pilot scale using a uniquely designed, patent-pending reactor bridging laboratory research with industrial practice. Experiments were conducted at 20 °C·min⁻¹, temperatures of 500–800 °C, and pressures from 0.1 to 1.0 MPa. PF pyrolysis produced mainly gas (70.1%), suitable for energy recovery, with smaller fractions of char (15.3%) and oil (14.6%). LG yielded predominantly char (59.9%), with lower gas (32.4%) and oil (7.7%) outputs. Co-pyrolysis revealed limited synergistic effects. Rising temperature promoted gas formation, reduced char, and improved its calorific value through carbon enrichment. Elevated pressure enhanced char yield and unexpectedly increased hydrogen content, suggesting complex thermochemical behaviour. The results confirm the scalability of laboratory findings and highlight pyrolysis as a practical WtE pathway for valorising protein-rich residues and low-rank coals, contributing to cleaner, more sustainable energy systems. Full article

While material extrusion via fused filament fabrication (FFF) offers design flexibility and rapid prototyping, its practical use in engineering is limited by mechanical challenges, including residual stresses, geometric distortions, and potential interlayer debonding. These issues arise from the dynamic thermal profiles during FFF, including temperature gradients, non-uniform hardening, and rapid thermal cycling, which lead to uneven internal stress development depending on fabrication parameters and object topology. These problems can compromise the structural integrity and mechanical properties of FFF parts, especially when the load-bearing capacity and geometric accuracy are critical. This study focuses on polylactic acid (PLA) due to its widespread application in engineering. It introduces a computational framework for coupled thermo-mechanical simulations of the FFF process using ABAQUS (Version 2020) finite element software. A key innovation is an automated subroutine that converts G-code into a time-resolved event series for finite element activation. The simulation framework explicitly models the sequential stages of printing, cooling, and detachment, enabling prediction of adhesive loss and post-process warpage. A transient thermal model evaluates the temperature distribution during FFF, providing boundary conditions for a mechanical simulation that predicts residual stresses and warping. Uniquely, the proposed model incorporates the detachment stage, enabling a more realistic and experimentally validated prediction of warpage and residual stress release in FFF-fabricated components. Although the average deviation between predicted and measured displacements is about 10.6%, the simulation adequately reflects the spatial distribution and magnitude of warpage, confirming its practical usefulness for process optimization and design validation. Full article

Background: Fixed-duration venetoclax plus rituximab (VR) is a standard therapy for relapsed/refractory (R/R) chronic lymphocytic leukemia (CLL). However, evidence supporting its use after covalent BTK inhibitor (c-BTKi) therapy is scarce in clinical trials and limited in real-world settings. Objectives: To assess the efficacy and safety of VR in a real-world cohort of patients with R/R CLL, including cBTKi-pretreated individuals, and to contextualize outcomes alongside published real-world studies and registrational trials of alternative therapies. Methods: We retrospectively analyzed 37 patients with R/R CLL treated with VR at our center between April 2018 and November 2024. Baseline characteristics, treatment responses, minimal residual disease (MRD), and adverse events were recorded. Survival was estimated using the Kaplan–Meier method. A structured review of relevant real-world evidence and pirtobrutinib clinical trials was also conducted. Results: Median age was 67 years; 35.1% had prior cBTKi exposure. The overall response rate (ORR) was 91.7% (22/24 evaluable patients), with 66.7% achieving complete remission (CR). Among evaluable c-BTKi-pretreated patients, the ORR was 87.5% (7/8) and the CR rate was 62.5%. Undetectable MRD (uMRD) rates were 78.6% in peripheral blood and 71.4% in bone marrow. Thirty-month progression-free survival (PFS), time to next treatment (TTNT), and overall survival (OS) were >90% for the whole cohort and for c-BTKi-pretreated patients. The most frequent adverse event was neutropenia grade ≥ 3, especially during combination therapy, which is easily managed with GCSF support. Conclusions: Our real-world evidence shows that VR is an effective and well-tolerated option even after c-BTKi therapy in R/R CLL. These data complement evidence from emerging therapies and inform post-c-BTKi treatment selection in clinical practice. Full article

Plastic waste and industrial residues like steel slag pose significant environmental challenges, with limited recycling solutions. This study investigates a sustainable approach by repurposing waste polystyrene and steel slag into composite membranes via electrospinning for membrane distillation applications. Steel slag incorporation enhanced membrane porosity, hydrophobicity, and thermal stability, with process optimization performed through response surface methodology by varying slag content (0–10 wt%), voltage (15–30 kV), and feed rate (0.18–10 mL·h⁻¹). Optimized membranes achieved a reduced fiber diameter (1.172 µm), high porosity (82.3%), and superior hydrophobicity (contact angle 102.2°). Mechanical performance improved with a 12% increase in tensile strength and a threefold rise in liquid entry pressure over pure polystyrene membranes, indicating greater durability and wetting resistance. In direct contact membrane distillation, water flux improved by 15% while maintaining salt rejection above 98%. Under photothermal membrane distillation, evaporation rates rose by 69% and solar-to-thermal conversion efficiency by 60% compared to standard PVDF membranes. These results demonstrate the feasibility of transforming waste materials into high-performance, durable membranes, offering a scalable and eco-friendly solution for sustainable desalination. Full article

Multimedia Forensics (MMF) investigates techniques to automatically assess the integrity of multimedia content, e.g., images, videos, or audio clips. Data-driven methodologies like Neural Networks (NNs) represent the state of the art in the field. Despite their efficacy, NNs are often considered “black boxes” due to their lack of transparency, which limits their usage in critical applications. In this work, we assess the interpretability properties of Deep High-Frequency Residuals (DHFRs), i.e., noise residuals extracted from images by NNs for forensic purposes, that nowadays represent a powerful tool for image splicing localization. Our research demonstrates that DHFRs not only serve as a visual aid in identifying manipulated regions in the image but also reveal the nature of the editing techniques applied to tamper with the sample under analysis. Through extensive experimentation on spliced amplitude Synthetic Aperture Radar (SAR) images, we establish a correlation between the appearance of the DHFRs in the tampered-with zones and their high-frequency energy content. Our findings suggest that, despite the deep learning nature of DHFRs, they possess significant interpretability properties, encouraging further exploration in other forensic applications. Full article

Background: The treatment of methicillin-resistant Staphylococcus aureus (MRSA) infections is extremely challenging due to its antibiotic resistance, and the combination of plant active ingredients with antibiotics represents a potential strategy to address this issue. Methods: We determined the combinatorial relationship between baicalin (BA) and kanamycin (KM) using the checkerboard dilution method. The antibacterial activity of the baicalin–kanamycin (BA/KM) combination was evaluated through growth curve determination assays and scanning electron microscopy (SEM). The effects of the BA/KM combination on the cell membrane and cell wall of MRSA were analyzed using reactive oxygen species (ROS) detection assays, intracellular protein leakage experiments, alkaline phosphatase (AKP) activity assays, laser scanning confocal microscopy (LSCM) observations, and molecular docking simulations. The antibiofilm activity and related mechanisms of the BA/KM combination were elucidated via crystal violet staining, MTT assay, phenol-sulfuric acid method, congo red staining, staphyloxanthin determination assays, and quantitative real-time polymerase chain reaction (qPCR). The safety of the BA/KM combination was assessed through hemolytic activity analysis, and its anti-MRSA efficacy was evaluated on lettuce. Results: BA/KM combination showed a synergistic antibacterial effect on MRSA USA300. Mechanistic studies revealed that BA may interact with amino acid residues of peptidoglycan synthetase PBP2a to hinder peptidoglycan synthesis, thereby facilitating KM penetration through the cell wall. Subsequently, BA binds to amino acid residues of the membrane transporter NorA, leading to disruption of cell membrane homeostasis and enhancing KM’s ability to induce intracellular ROS accumulation in MRSA. Furthermore, the BA/KM combination reduced MRSA biofilm formation by 77.85% and decreased the metabolic activity of biofilm cells by 42.93% through inhibiting the synthesis of biofilm components EPS and PIA. Additionally, this combination suppressed the synthesis of staphyloxanthin and downregulated the expression of agrA and agrC genes. When 1/8 MIC BA was combined with 1/4 MIC KM, the count of MRSA on lettuce surfaces was reduced by 0.88 log CFU/cm², an effect comparable to that of 0.2% (v/v) hydrogen peroxide. Conclusions: According to these findings, the BA/KM combination may offer a promising option for enhancing antibacterial efficacy through synergism, reducing antibiotic usage concentrations, and limiting MRSA transmission in fresh agricultural products. Full article

Environmental stressors, particularly agricultural pesticides, can influence both growth and physiology in developing birds, yet the relationship between morphometric condition indices and biochemical biomarkers remains poorly understood. We investigated body mass, beak length, tarsus length, and body condition index (BCI) alongside plasma and S9 biomarkers, including the activities of acetylcholinesterase (AChE), carboxylesterase (CES), and glutathione S-transferase (GST), as well as the levels of glutathione (GSH) and reactive oxygen species (ROS) in nestling White Storks (Ciconia ciconia) from Croatia. Bayesian undirected graphical model (BUGM) inferred a disconnected correlation structure composed of two communities, with a strong beak length–GSH association. Biomarkers further exhibited plasma-specific affinity: plasma markers reflected short-term adjustments, whereas S9 enzymes represented distinct metabolic pathways. Overall, morphometry and physiological status showed only limited integration, restricted mainly to plasma biomarkers, and residual body condition index did not serve as a reliable proxy for physiological stress. We conclude that integrated monitoring approaches, combining morphometric and biochemical profiling, provide a more nuanced assessment of nestling condition and strengthen the use of White Storks as sentinels of agroecosystem health. Full article

In the task of integrated circuit micrograph acquisition, image super-resolution reconstruction technology can significantly enhance acquisition efficiency. With the advancement of deep learning techniques, the performance of image super-resolution reconstruction networks has improved markedly, but their demand for inference device memory has also increased substantially, greatly limiting their practical application in engineering and deployment on resource-constrained devices. Against this backdrop, we designed image super-resolution reconstruction networks based on feature reuse and structural reparameterization techniques, ensuring that the networks maintain reconstruction performance while being more suitable for deployment in resource-limited environments. Traditional image super-resolution reconstruction networks often redundantly compute similar features through standard convolution operations, leading to significant computational resource wastage. By employing low-cost operations, we replaced some redundant features with those generated from the inherent characteristics of the image and designed a reparameterization layer using structural reparameterization techniques. Building upon local feature fusion and local residual learning, we developed two efficient deep feature extraction modules, and forming the image super-resolution reconstruction networks. Compared to performance-oriented image super-resolution reconstruction networks (e.g., DRCT), our network reduces algorithm parameters by 84.5% and shortens inference time by 49.8%. In comparison with lightweight image reconstruction algorithms, our method improves the mean structural similarity index by 3.24%. Experimental results demonstrate that the image super-resolution reconstruction network based on feature reuse and structural reparameterization achieves an excellent balance between network performance and complexity. Full article