The phenomenon of Balkan high-mountain fen plant endemism, unique within Europe, remains poorly studied. We compiled an original dataset of wetland vegetation-plot data across nine countries where Balkan endemics and subendemics occur. We tested the diversity patterns for endemics, non-endemic habitat specialists of bryophytes and vascular plants and matrix-derived taxa with respect to climate, palaeoclimate, edaphic and topographic factors. We used Spearman correlations, Spatial Autoregressive Error Models, Canonical Correspondence Analysis and Principal Component Analysis of abiotic factors with richness patterns projected using Generalised Additive Models. Two major hotspots of fen endemism, centred around the Rila and Korab Mts, were identified. Species composition of endemics generally follows an east-to-west gradient that coincides with palaeoclimatic patterns. Palaeoclimate was influential for community-scale richness of bryophytes and narrow-range endemics, the latter being positively affected by the Bølling-Allerød precipitation peak and the wet-tundra climate during the glacial. The richness of other groups was affected by recent climate, edaphic factors and topography. Water pH differentiated major vegetation types and affected the richness of matrix-derived taxa and fen bryophytes, but had a minor impact on Balkan endemics, except those associated with serpentinite. Several endemic taxa usually co-exist even in small wetland patches. Palaeoclimate shows that fen endemics evidently experienced drier, warmer summers than today over the last 21,000 years; yet they were more affected by past precipitation. If we conserve Balkan high-mountain fens, their catchment properties and hydrology, and maintain moderate grazing, we increase the likelihood that their endemics will withstand ongoing climate change.

Organ-targeted nanomedicine is often defined by anatomical destination, yet delivery performance is determined by the biological barriers, disease-associated remodeling, and nano-bio interface interactions encountered before therapeutically relevant cells are reached. Whole-organ accumulation, tissue fluorescence, or bulk biodistribution can demonstrate tissue arrival, but stronger targeting claims require evidence matched to the intended delivery task, including barrier crossing or penetration, spatial localization, target-cell exposure, cargo release, target engagement, functional activity, and safety. We introduce the Barrier-to-Design Codes framework, a conceptual approach that links measurable barrier phenotypes to conditional material and nano-bio interface parameters, claim-matched validation evidence, failure boundaries, and translational gates. The framework complements existing reporting, delivery-system design, and translational frameworks by treating the barrier phenotype-rather than the organ label or material class-as the primary unit of analysis. It integrates five linked domains: barrier architecture, pathological remodeling, acquired nano-bio interface behavior, including protein adsorption and immune recognition, validation, and translation. Using the blood-brain barrier, lymph nodes and immune organs, liver and metabolic organs, skin and mucosal barriers, and cartilage/extracellular-matrix niches as representative systems, we examine how particle size, charge, stiffness, ligand density, coating, release, protein adsorption, and immune recognition can acquire different functional meanings across barrier contexts. The framework separates transferable reasoning principles from non-transferable platform assumptions and requires organ-level signals to be resolved through barrier-matched models, spatial and cell-resolved exposure, pharmacokinetic/pharmacodynamic interpretation, functional response, manufacturing control, and repeat-dose safety. The Barrier-to-Design Codes framework is not a nanomaterial taxonomy, scoring system, or universal design prescription. Its intended value lies in reducing unsupported targeting claims, improving alignment between design hypotheses and validation evidence, and supporting more interpretable and translation-aware decisions, while remaining open to prospective empirical testing and refinement.

Citizen reporting of biosecurity threats is an important yet inconsistently practiced element of environmental surveillance. This study examines the behavioural factors shaping reporting in Aotearoa New Zealand by conceptualising reporting as a multi-step process influenced by psychological, social, and structural conditions. Drawing on behavioural science, we apply the Capability, Opportunity, Motivation-Behaviour model alongside audience segmentation to identify barriers and drivers of reporting using a national survey of 957 New Zealand adults. Latent profile analysis identified four distinct behavioural segments-Engaged Reporters, Reluctant Reporters, Willing but Unaware and Disconnected Non-reporters-characterised by differing levels of capability, opportunity and motivation. While public awareness of the importance of biosecurity has grown, personal engagement in surveillance activities remains comparatively limited. These findings suggest that constraints to reporting arise from a combination of informational, motivational and structural factors, which vary across population groups. Rather than prescribing specific interventions, the study provides a behavioural diagnostic that can inform the development of more targeted and context-sensitive engagement strategies. Importantly, the effectiveness of citizen reporting depends not only on individual behaviour but also on the capacity and responsiveness of the systems that receive and act on reports. As such, efforts to enhance reporting should consider both citizen engagement and system-level design. By integrating behavioural theory with audience segmentation, this study contributes to a more nuanced understanding of citizen participation in biosecurity and highlights the importance of aligning behavioural insights with practical and institutional considerations. Future research should evaluate the effectiveness of intervention strategies and examine how reporting behaviours translate into biosecurity outcomes in real-world contexts.

Accurate classification of brain tumors from MRI scans requires models capable of capturing both fine-grained structural details and broader contextual patterns. This study proposes a hybrid Wavelet-EfficientNetV2-Vision Transformer (ViT) framework that integrates multi-resolution frequency enhancement, hierarchical spatial encoding, and global dependency modeling into a unified representation. The Discrete Wavelet Transform (DWT) enriches directional and boundary-aware features, EfficientNetV2 extracts robust localized patterns, and the ViT captures long-range contextual relationships essential for differentiating tumor categories with overlapping visual characteristics. The method was evaluated on the Kaggle-MRI dataset consisting of four classes: glioma, meningioma, pituitary tumor, and no tumor. Experimental results show that the proposed model achieves 98.25% accuracy, with consistent gains in precision, recall, and F1-score compared to standalone CNNs, transformer models, and recent state-of-the-art approaches. Additional analyses, including confusion matrices, ROC-AUC evaluation, and stratified five-fold cross-validation, demonstrate strong class-level stability and generalization. These findings indicate that combining frequency, spatial, and contextual information provides a more discriminative feature space for MRI-based tumor classification. The proposed framework offers a promising direction for developing reliable and high-performance computer-aided diagnosis systems, with future work aimed at multi-center validation and extension to volumetric (3D) modeling.

Long-term mine drainage represents one of the most intensive anthropogenic stresses imposed on groundwater systems, yet its system-scale hydrological impacts in multi-aquifer settings remain insufficiently understood. In this study, we investigate the response of a regional, vertically layered groundwater system to approximately 50 years of sustained mine drainage, with particular emphasis on the contrasting effects of single-well and group-well pumping. A physically based transient groundwater flow model was developed for a coal-bearing sedimentary basin and applied to simulate baseline conditions, staged single-mine drainage, and multi-mine group drainage scenarios. The results show that prolonged drainage induces a systematic reorganization of groundwater flow fields that extends well beyond localized drawdown around mining panels. Groundwater system behavior exhibits a clear transition from an initial storage-dominated response toward a boundary-controlled and vertically integrated flow regime, driven by enhanced lateral recharge and inter-aquifer leakage under sustained pumping. Importantly, groundwater responses under group-well pumping are inherently non-linear and cannot be approximated as a linear superposition of single-well effects. Interaction and overlap of capture zones among multiple drainage centers redistribute hydraulic gradients and produce pronounced spatial heterogeneity in groundwater-level responses. Aquifer stratification and vertical connectivity are shown to play a critical role in controlling the magnitude, spatial extent, and persistence of drainage-induced impacts. Although derived from a mining context, the identified hydrological mechanisms are applicable to other groundwater systems subjected to sustained, spatially clustered extraction. The findings highlight the need to treat long-term, clustered pumping as a system-scale hydraulic forcing in groundwater impact assessment and management.

We aim to (1) explore the main barriers to treatment faced by lung cancer patients, from a provider's perspective, and (2) the existing support systems in addition to recommended mitigation strategies to enhance the preparedness process. Our study invol ves 50 surveys and 10 interview responses, in a mixed-method design. We conducted descriptive statistics and thematic analysis through a hybrid inductive-deductive approach. We based our analysis on the Optimal Care Pathways framework and Andersen's Behavioral Model. The doctors identified challenges to patient readiness for treatment, such as a lack of knowledge about treatment options, emotional load, logistical barriers (e.g., transportation), and financial pressure. The providers mentioned some initiatives adopted to try to address these challenges, namely in-person instruction, printed materials, and transportation services provided by social workers. However, they emphasized the potential role of accessible digital tools to complement existing educational and support processes. Our results highlight the need to create educational protocols that are standardized yet phase-specific to address the challenges of each treatment phase. These suggestions may improve overall therapy results, decrease delays, and increase patient adherence.

Recombinant human type III collagen (rhCol III) is a non-animal-derived, biomimetic protein with notable regenerative potential. Nevertheless, its therapeutic efficacy in chronic inflammatory skin conditions such as atopic dermatitis (AD) remains insufficiently explored. This study aimed to evaluate the efficacy and safety of rhCol III as a topical therapeutic agent for AD. Utilizing an ovalbumin (OVA)-induced murine model of AD, topical application of rhCol III substantially ameliorated skin lesions, reduced scratching behavior and epidermal thickening, and suppressed mast cell infiltration and Th2 cytokines (IL-4, IL-13, and TSLP), while also demonstrating a decreasing trend in serum IgE levels. In vitro assays demonstrated that rhCol III inhibited the expression of IL-6, IL-1β, CXCL17 and CXCL22 in TNF-α and IFN-γ-stimulated HaCaT cells, confirming its anti-inflammatory properties. Long-term safety assessments revealed no structural abnormalities in the skin or major organs and no clinically relevant alterations in hematological parameters, bone mineral content (BMC), fat mass, or lean mass following 120 days of continuous topical application in mice. Preliminary clinical observations in infant patients with AD indicated that rhCol III was well tolerated, with observable symptom improvement, particularly in mild-to-moderate cases where topical rhCol III alone produced discernible therapeutic benefits. Collectively, these findings support the therapeutic potential and safety of rhCol III as a novel non-steroidal treatment strategy for AD.

This study evaluates the development and performance of convolutional neural network (CNN) models for classifying subvisible particles in parenteral products using a comprehensive and pharmaceutically relevant particle image library. Specifically, representative morphologies of glass lamellae were curated for the first time and incorporated into the particle library, along with other intrinsic particle types such as protein aggregates, silicone oil droplets, and rubber. CNN models trained on complete particle libraries achieved high accuracy and stable classification across complex multi particulate mixtures, whereas models trained on limited categories or incomplete morphologies showed reduced robustness and increased misclassification, particularly for transparent or low‑contrast particles. These findings demonstrate that classification performance is highly dependent on the representativeness of the training dataset and the inclusion of challenging particle classes. The results also align with expectations outlined in compendial documents and regulatory guidance, highlighting the importance of well‑curated particle libraries and explainable model behavior to support reliable, scientifically justified particle identification in pharmaceutical quality assessments of parenteral products.

Three-dimensional (3D) gastric organoids have become valuable experimental models for gastric cancer research, as they closely mirror tumour architecture, molecular features, and treatment responses. Unlike two-dimensional cultures and animal models, gastric organoids capture key physiological and pathological characteristics of the human stomach, allowing more meaningful investigation of tumour development, host-microbe interactions, and epithelial-stromal crosstalk. These systems retain the genetic and phenotypic traits of their tissue of origin, making them suitable for both mechanistic studies and long-term experimentation. This review outlines the development of gastric organoid technology, from early advances in adult stem cell-derived cultures to recent methodological refinements that have expanded their experimental and clinical relevance. We discuss how organoid models have helped address the challenge posed by the marked molecular and histological heterogeneity of gastric cancer, which continues to complicate diagnosis and therapeutics. In this regard, patient-derived gastric organoids and associated biobanks have shown strong potential as predictive tools for drug response, as they reliably preserve tumour structure and gene expression patterns observed in primary tissues. In addition, we highlight recent improvements in culture matrices and experimental approaches that have enhanced organoid robustness and reproducibility. The growing use of gastric organoids as preclinical platforms for drug screening and the study of cancer-related signaling pathways reflects the rapid maturation of the field. We consider current limitations, technical challenges, and future directions for gastric organoid research, emphasizing their role as complementary models that bridge the gap between conventional in vitro systems and in vivo studies in gastric cancer.

The Landing Error Scoring System (LESS) is an assessment tool used for identifying movement patterns linked with non-contact anterior cruciate ligament injuries during a double-leg jump-landing; however, the LESS is scored by experts using 2D video recordings, which limits large-scale screening. This study explores the viability of using inertial measurement unit (IMU) data to automate scoring of 17 LESS items, using conventional machine learning model architecture. Forty healthy participants completed six jumps each, and raw movement data from three IMU sensors placed on the sacrum and medial-inferior aspects of the tibias were processed and segmented into the key phases of the double-leg jump-landing. A total of 218 jumps were used to train supervised machine learning models using various subsets of the processed dataset, including extracted temporal and statistical features. Performance metrics were extracted, and the best performing models for each scoring item were evaluated against a majority-class baseline classifier (ZeroR). Results indicated moderate improvements over ZeroR, ranging from 0.3 % to 22.5 %, with some models demonstrating limited recognition of the minority class. These findings provide a foundation for future research in IMU-based LESS automation, and improving the accessibility of movement screening tools.