Aortic valve calcification refers to the accumulation of calcium on the aortic valve leaflets, which can lead to aortic stenosis. A deeper understanding of the hemodynamic implications of altered valve properties is necessary. Consequently, it is essential to explore the biomechanical characteristics of aortic valve leaflets that are prone to calcification. In order to analyze fluid dynamics in an aortic segment with leaflets exhibiting varying stiffness, a two-way fluid–structure interaction model was developed. The behavior of the leaflets was simulated using two constitutive laws: linear elastic and isotropic hyperelastic, followed by numerical evaluations and comparative studies. The hyperelastic model was assessed using material parameter values within the ranges of 22–60 and 22–60 kPa, respectively. Young's modulus of the valve leaflets varied from 1 to 22 MPa, while Poisson's ratio was between 0.35 and 0.45. A strong correlation was observed between Poisson's ratio and wall shear stress. With an elastic modulus of 22 MPa and the maximum Poisson's ratio of 0.45, the peak wall shear stress reached 81.78 Pa during maximum flow velocity and full valve opening, whereas the minimum wall shear stress was recorded at 0.38 Pa. From the findings of this study, it can be concluded that when accounting for the isotropic structure and nonlinear properties of valve leaflets, the Delfino hyperelastic model provides a more precise representation of their intricate behavior.

The authors would like to acknowledge the funding provided by the Ministry of Science and Higher Education of the Russian Federation (Project № FSNM-2024-0009).

Precise brain tumor segmentation from MRI scans is critical for clinical diagnosis, surgical planning, and treatment monitoring. However, determining an optimal deep learning architecture remains a significant challenge due to the high-dimensional hyperparameter space, variability in tumor morphology, and differences across MRI modalities. This paper presents a novel approach that integrates Bayesian Optimization (BO) to systematically tune the U-Net architecture for enhanced brain tumor segmentation performance. The proposed BO-UNet framework explores various encoder, bottleneck, and decoder configurations using a Gaussian Process-based surrogate model. The optimization is guided by a composite fitness function that averages the Dice Similarity Coefficient (DSC) and Jaccard Index (JI), ensuring accurate spatial overlap between predicted and ground truth masks. Experiments were conducted on two benchmark datasets: the Figshare Brain Tumor Segmentation (FBTS) dataset and the BraTS 2021 dataset, with a focus on Whole Tumor (WT) segmentation. The best architecture discovered—[64, 64, 64, 256, 64, 128, 256]—demonstrated superior performance. On the FBTS dataset, it achieved 0.9503 DSC and 0.9054 JI, while on BraTS 2021 it reached 0.9261 DSC and 0.8631 JI, outperforming several state-of-the-art deep learning models. Convergence plots and segmentation map visualizations illustrate the effectiveness of BO in guiding architectural evolution. These findings highlight the potential of data-driven optimization strategies for automatic model design in medical image analysis, particularly in domains requiring high precision and structural sensitivity.

Federated Learning (FL) offers a promising paradigm for privacy-preserving, collaborative machine learning; however, the presence of non-independent and identically distributed (non-IID) data among clients significantly affects global model performance. This research proposes a novel architecture that combines Knowledge Distillation (KD) with Vision Transformer (ViT) models and hierarchical fully homomorphic encryption (FHE) to address both the non-IID data challenge and privacy preservation in FL. The proposed framework employs an aggregator server to homomorphically aggregate encrypted local model parameters, which are then decrypted and averaged by a separate federated server, ensuring that only clients retain access to their unencrypted parameters. Traditional approaches that modify aggregation algorithms are computationally prohibitive or incompatible with FHE; in contrast, KD facilitates robust model adaptation to local client distributions, supports heterogeneous client architectures, and integrates seamlessly with encrypted workflows. Experimental results with two clients, one utilizing the CIFAR-10 dataset and another utilizing the Pascal VOC 2007 dataset (sharing common classes), demonstrate the efficacy of the approach. EfficientNet was used for local training with Pyfhel-based FHE applied to model parameter exchange. Without knowledge distillation, the system obtained an AUC of 0.78, which improved to 0.84 when applying ViT-based knowledge distillation. The findings highlight the proposed method's potential to enhance FL robustness, adaptability, and privacy, representing a viable and scalable solution for privacy-preserving collaborative learning in heterogeneous environments.

Fermentation is recognized as one of the oldest techniques for preserving and processing food. Lacto-fermented vegetables are increasingly valued as functional foods due to their enrichment with bioactive metabolites released during fermentation. While the solid vegetable fraction is widely studied, fermentation brine remains an underutilized resource despite its potential as a source of health-enhancing compounds. This study aims to examine the temporal changes in antioxidant capacities of fermentation brines.

The antioxidant potential of brines from fermented vegetables samples prepared from different combinations of cabbage, beetroot, carrot, cucumber, and bell pepper was evaluated during fermentation. Antioxidant activity was measured at different time points (0, 12, 24, 72, 120, 240, 360, and 504 hours) using four in vitro tests: ABTS, DPPH, FRAP, and TAC.

All brine samples showed antioxidant activity, which varied significantly over time. ABTS showed a strong radical scavenging activity, reaching up to 96% on day 21, while DPPH activity was moderate, with a maximum of 54% the same sampling time. The FRAP activity highest absorbance value reached 2.071 on day 21, while TAC peaked at an absorbance value of 1.788 on the same day, confirming the antioxidant potential of brines. These variations can be explained by the plant matrix and the gradual diffusion of bioactive compounds during fermentation. Overall, this study highlights fermentation brines as a promising and lasting source of natural antioxidants. Rather than being perceived as waste, these products could be used as useful ingredients for nutritional and food applications, contributing to sustainable and circular food systems.

Medicinal plants are precious reservoirs of bioactive molecules; they are able to produce diverse natural substances. Currently, research is focused on creating new applications and taking advantage of the qualities of these resources in many industries, including the food, cosmetics, and pharmaceutical industries [1]. In line with this global interest in medicinal plants In this work we explore the antimicrobial properties of the Eriobotrya japonica, the extract has been obtained after plant leaves were meticulously collected during their optimal growth period, followed by a careful process of drying and grinding. Employing the Maceration technique, we utilized ethanol as the solvent for extraction, with subsequent defatting accomplished using hexane. The total phenolic content in the Eriobotrya japonica extract was determined to be 384.19 µg/mL gallic acid equivalents, indicating that the plant is rich in phenolic compounds. The antioxidant activity also was evaluated by various antioxidant methods, including 2,2-diphenyl-1-picrylhydrazyl (DPPH), 2,2-Azinobis (3-ethylbenzothiazolin) -6-sulphonic acid (ABTS), Ferric Reducing Antioxidant Power (FRAP), and phenanthroline. Those various antioxidant activities were compared to standard antioxidants such as butylated hydroxyanisole (BHA) and butylated hydroxytoluene (BHT). The extract demonstrated strong antioxidant activity across all methods, with low IC₅₀ values 253.6 , 1.68 , 32.68 and 4.11 µg/mL respectively, Overall, the results suggest that Eriobotrya japonica is a valuable plant with strong antioxidant properties, making it a potential candidate for various health-related applications.

Malaria remains a critical public health burden in sub-Saharan Africa, contributing significantly to illness and mortality, particularly among children and pregnant women. Rapid and accurate diagnosis is essential for timely treatment and effective disease management. Convolutional Neural Networks (CNNs) have shown substantial promise in automating malaria detection from microscopic blood smear images, but their performance heavily depends on optimal hyperparameter tuning, a task that is computationally intensive and highly sensitive to initial conditions. To address this challenge, this study proposes an enhanced Adaptive Marine Predators Algorithm (AMPA) for efficient hyperparameter optimization. The proposed method introduces a dynamic step-size adjustment strategy, which adaptively modifies the search behavior in response to real-time validation loss trends during training. This mechanism improves convergence stability and helps the optimizer focus on promising regions of the search space. Furthermore, a multi-objective fitness function is employed to jointly optimize classification accuracy, generalization capability, and computational efficiency. The effectiveness of the proposed approach is demonstrated using the publicly available Kaggle Malaria Cell Images Dataset, which consists of over 27,000 annotated images of parasitized and uninfected red blood cells. Empirical results show that the adaptive MPA consistently outperforms conventional optimization strategies, yielding CNN configurations with superior detection accuracy and faster convergence. These findings highlight the potential of intelligent, nature-inspired optimization algorithms in improving the deployment of deep learning-based diagnostic systems in real-world, resource-constrained healthcare settings, and contribute to the broader goal of enhancing malaria control through automated, scalable diagnostic tools.

Agricultural productivity is influenced by a complex interplay of environmental conditions, soil characteristics, and farm management practices. Traditional farming methods often lack the precision and adaptability required to optimize these dynamic variables, limiting crop yield potential and sustainability. In response to this challenge, this study presents an intelligent crop optimization framework that leverages the capabilities of bio-inspired metaheuristic algorithms, Particle Swarm Optimization (PSO) and Ant Colony Optimization (ACO). Inspired by natural phenomena such as bird flocking and ant foraging, PSO and ACO are employed to explore optimal combinations of key agricultural parameters, including soil nutrient composition, pH levels, rainfall, and temperature. A neural network is trained to evaluate these parameter configurations, providing a performance-guided feedback mechanism that directs the search toward high-yield solutions. By integrating PSO and ACO with a neural predictive model, the proposed hybrid system combines the global search power of evolutionary algorithms with the pattern recognition strength of deep learning. This synergy enhances both the accuracy and robustness of decision-making in agricultural settings. The model not only adapts to changing environmental inputs but also supports real-time optimization, making it highly suitable for precision agriculture applications. Experimental results demonstrate that the system can effectively recommend parameter configurations that maximize yield while maintaining resource efficiency. The proposed approach offers a scalable, data-driven solution that empowers farmers with intelligent tools for informed and sustainable agricultural planning. This study contributes a novel and adaptive computational framework for optimizing crop yields, bridging the gap between artificial intelligence and modern farming practices.

Introduction:
Bioengineered mesenchymal stem cell (MSC)-based 3D constructs hold immense promise for regenerative medicine and biomedical research. Realizing their full potential necessitates effective preservation methods. While cryopreservation is the current gold standard, it presents challenges such as reduced viability and demanding cold chain logistics. This study investigates novel approaches for ambient storage of MSCs in spheroids, alginate microspheres (AMSs), and macroporous scaffolds, aiming to develop effective technology for short-term storage.

Methods:
Human adipose tissue-derived MSCs were used. Spheroids were formed using the "hanging drop" method, AMSs via electrospraying, and scaffolds by means of plasma cryogelation before cell seeding. Constructs were cultured for 3 days and then stored in complete medium at 22 °C. Viability/apoptosis (6-CFDA/annexin V-Cy3), metabolic activity (resazurin), differentiation potential, and reactive oxygen species (ROS) levels (DCFH-DA) were assessed before and after storage.

Results and Discussion:
Ambient storage of MSCs in suspension resulted in significant viability loss. Cells within all 3D constructs demonstrated preserved viability, metabolic activity, and differentiation ability for up to 7 days of storage. Basal metabolic activity was decreased in spheroids and AMSs, and these constructs maintained unchanged levels of annexin-positive cells and ROS throughout storage. Annexin-positive cell number increased minimally in scaffolds and notably in suspension, with ROS rise in suspension after 7-day storage.

Conclusions:
This study demonstrates the feasibility of ambient storage for MSC-based 3D constructs and represents a significant step towards developing a safer, cost-effective, cold chain-independent solution for their short-term storage and transportation.

This study was supported by the National Research Foundation of Ukraine (project № 2021.01/0276).

The study of the Vegetation is a crucial factor in the ecosystem. This study investigates the improved classification for urban vegetation on Giglio Island (in Italia) by integrating data from Sentinel-1 (Synthetic Aperture Radar) and Sentinel-2 (Multispectral) from the European Space Agency's Sentinel constellation. The analysis leverages the power of Random Forest Algorithm and Sklearn python’s library to create a classification map. Usually, urban vegetation mapping relies on a single data source resulting in limitations in accuracy and the abilities to differentiate vegetation types. The vegetation treated were: Mediterranean macchia, grasslands, coastal vegetation, pine forest. First, we performed independent classification vegetation using Normalized Difference Vegetation Index and Radar Vegetation Index. Secondly, to enhance classification accuracy, by incorporating a combination of each index with the Modified Normalized Difference Water Index and Soil Adjusted Vegetation Index we identified the vegetation classes exhibiting the highest prevalence. Thirdly using the proposed fusion method we combined the Modified Normalized Difference Water Index and Soil Adjusted Vegetation Index with the fusion between Normalized Difference Vegetation Index and Radar Vegetation Index and we compared the density variations among vegetation types. The results showed the effectiveness of the fusion approach significantly improved the classification accuracy, a perfect 100% of overall accuracy and Kappa index in all prediction elements. It showed a classification result by class for Random Forest Algorithm where each vegetation types were “present”, and showed the agreement level by class for Random Forest Algorithm some vegetation types were perfect, moderate, fair, slight, poor, very poor, inexistent.

The rapid growth of digital text highlights the need for effective summarization. Traditional graph based methods, like TextRank, often fall short by relying primarily on lexical similarity, which can miss crucial semantic connections and deeper contextual meaning. This study proposes a two stage summarization framework that integrates an enhanced graph-based ranking mechanism with a metaheuristic optimization strategy. In the initial phase, we modified the conventional TextRank algorithm by redefining edge weights through a combination of lexical, structural, and semantic attributes, specifically sentence position, bigram overlap, and SBERT based semantic similarity. This multi feature integration enhances the estimation of sentence significance by effectively capturing both surface level and contextual relationships. In the subsequent phase, we present a refined Snake Optimization Algorithm that identifies optimal subset of sentences through the application of a fitness function. This function integrates ROUGE-1, ROUGE-2, ROUGE-L metrics, SBERT-based semantic similarity aligned with the reference summary, as well as a sentence threshold penalty to control redundancy and length. Findings on the Medium Article datasets demonstrate improved summarization quality in terms of both lexical and semantic metrics, validating the effectiveness of the proposed two stage strategy. This research significantly contributes to the advancement of extractive summarization models by combining graph based ranking with semantically informed optimization.