Abstract: Background/Objectives: Advanced technologies in type 1 diabetes mellitus (T1DM) management have reshaped the strategies used to achieve optimal glucose control. Continuous Subcutaneous Insulin Infusion (CSII) and Automated Insulin Delivery (AID) systems are effective alternatives to multiple daily injections (MDI). This study aims to evaluate glycemic regulation in children and adolescents transitioning from MDI to in-sulin pumps and to raise awareness among patients and their families regarding the benefits of these systems. Methods: 50 pediatric patients with T1DM (24 males, 26 fe-males; mean age 10.76 ± 3.2 years) were evaluated. Cycle 1 established MDI metrics 3 months pre-transition. In cycle 2, patients transitioned either to an AID system (Medtronic MiniMed 780G, 78%), or a non-automated system (Omnipod DASH, 22%). Data were assessed at 3- and 6-months post-initiation. Parameters assessed were: Glycosylated hemoglobin (HbA1c), Time In Range (TIR), Time Above Range (TAR), Time Below Range (TBR), Glucose Management Indicator (GMI), Coefficient of Variation (CV). Results: The cohort exhibited a statistically significant increase in TIR (p=0.0038) with mean values of 70.9% at 3 months and 70.8% at 6 months. TAR significantly reduced (p=0.033) to 26.5% and 24.3% at 3 and 6 months, respectively. Sub-analysis in the AID group, revealed a marked increase in TIR (p=0.0001) alongside significant reductions in TAR (P=0.0009) and GMI (p=0.03). Conclusion: Transitioning from MDI to insulin pump therapy, particularly AID systems, is transforming the clinical landscape of T1DM management. The con-sistency of these results across age groups indicates that AID systems can successfully overcome pediatric and adolescent diabetes management challenges.

Abstract: Regular graphs are classical symmetric structures in graph theory, where each vertex has identical degree and the overall topology often exhibits strong automorphism properties. However, practical systems frequently require heterogeneous constraints, which can be modeled by introducing vertex colorings and non-uniform degree requirements, leading to controlled symmetry breaking. In this paper, we investigate two-tone factors in edge-connected regular graphs and claw-free cubic graphs under arbitrary red-blue vertex colorings. Using the framework of parity (g,f)-factors, we establish two main existence results. First, we prove that every λ-edge connected r-regular graph admits a two-tone ({k},{k,k+2})-factor for any coloring, provided that r/λ⩽k⩽r−r/λ, k⩽r−2, and k|G| is even. Second, we show that every 3-edge connected claw-free cubic graph admits a two-tone ({0,1},{2,3})-factor regardless of the coloring configuration. Beyond existence, we provide a constructive algorithm by reducing the parity factor problem to an exact f-factor problem and further to a perfect matching problem via vertex-splitting techniques. We rigorously justify the correctness of this reduction and show that the desired factor can be computed in polynomial time. From a structural perspective, our results reveal that edge-connectivity serves as a stabilizing mechanism that preserves parity feasibility under arbitrary color-induced perturbations, while claw-free constraints enforce local density that prevents parity imbalance. This provides a symmetry-based interpretation of two-tone factors as a balance between global regularity and local asymmetry. These findings contribute to both the theoretical development of factor theory and its algorithmic realization, with potential implications for deterministic network design and resource allocation in structured systems.

Abstract: Galaxy classification is essential for understanding the formation and evolution of cosmic structures. However, faced with the explosive growth of astronomical observation data, traditional single-modality classification methods relying solely on spectroscopy or imaging have struggled to meet high-precision demands due to insufficient feature utilization and limited generalization capability. Therefore, multimodal fusion has emerged as a promising direction by leveraging information complementarity to overcome the limitations of single data sources. Accordingly, this paper proposes a model named Galaxy CosineNet (GCSNet), which integrates imaging, spectroscopic, and tabular data for high-precision galaxy classification. Specifically, the model employs dedicated encoders to process the three modalities separately and utilizes skip connections to preserve raw features. Furthermore, it incorporates a multi-head self-attention mechanism to deeply mine global cross-modal complementary information. Finally, these features are concatenated and fed into a cosine similarity classification head. Experimental results demonstrate that GCSNet achieves 97.15% accuracy in classifying star-forming, composite, active galactic nuclei (AGNs), and normal galaxies. This performance outperforms the best single-modal baseline, GaSNet, by 0.76% and mainstream multi-modal models such as MB-ISTL and the Transformer by over 1.6%. Consequently, the proposed GCSNet offers an effective and novel approach for research on automatic galaxy classification.

Abstract: Understanding the extent to which anthropogenic activity shapes vegetation dynamics is a central challenge in palaeoecology. In the Eastern Mediterranean, pollen-based studies have traditionally identified human impact through qualitative interpretations of anthropogenic indicators, particularly within the framework of the Beyşehir Occupation Phase (BOP) . However, quantitative comparison of anthropogenic signals across multiple sites remains limited. This study compiles pollen datasets from multiple lacustrine records across Anatolia (Türkiye) to construct a regional multi-site dataset and evaluates anthropogenic influence using a quantitative BOP period anthropogenic taxa integrated with Principal Component Analysis (PCA). Anthropogenic impact was quantified using a composite pollen index based on Olea, Juglans, Plantago lanceolata-type, Cerealia and Rumex acetosa-type taxa. The results reveal substantial spatial variability in anthropogenic signals, with combined pollen percentages ranging from less than 1% to 16% among lakes. PCA results show clear inter-site differentiation, with the first two components explaining 42.94% and 21.95% of the total variance, respectively. In particular Olea emerges as the most influential indicator, strongly contributing to the primary ecological gradient. These findings provide a quantitative extension of the traditionally qualitative BOP concept and demonstrate that anthropogenic influence is a fundamental and spatially heterogeneous component of vegetation dynamics across Anatolia. By integrating a composite anthropogenic index with multivariate analysis, this study offers a robust and transferable framework for comparing human–environment interactions across different regions and ecological settings.

Abstract: Accurate detection of special targets in unmanned aerial vehicle (UAV) remote sensing imagery under complex degradation conditions remains a critical challenge for intelligent surveillance systems. Existing detectors exhibit significant performance degradation when confronted with composite degradation factors such as blur, rain, snow, fog, low illumination, strong light, and electromagnetic interference. To address this limitation, we propose RHG-DETR (Riemannian Hyper-Graph Detection Transformer), a novel detection framework for robust special target detection under multi-type degradation in UAV remote sensing imagery. Using RT-DETR as the baseline, three synergistic innovations are introduced at the backbone, neck, and encoder levels. The Dynamic Receptive-field Hyper-graph Attention Network (DRHANet) replaces the conventional ResNet backbone, employing anisotropic dynamic depthwise separable convolution and a Riemannian Hyper-graph Fusion (RHGF) mechanism to model high-order semantic topology dependencies among target components. The Bi-directional Weighted Adaptive Fusion Network (BWAFN) constructs a two-stage bidirectional feature pyramid with learnable scale contribution weights and a lightweight spatial compensation upsampler to maintain cross-scale semantic consistency under atmospheric degradation. The Adaptive Sparse Multi-scale Encoder with Dynamic normalization (ASMED) reconstructs the AIFI encoder module by introducing sparse window self-attention to suppress background interference, a spatial-gated feedforward fusion to preserve geometric topology constraints of target sub-components, and coordinated dynamic normalization modules to stabilize encoding under extreme illumination and electromagnetic interference. On a self-constructed special target dataset comprising tanks, multiple launch rocket systems, and soldiers under seven degradation types, RHG-DETR achieves an mAP50 of 78.5%, surpassing the RT-DETR baseline by 3.7%, while reducing GFLOPs and parameter count by 34.4% and 28.8%, respectively, at an inference speed of 84.2 FPS. Consistent improvements on VisDrone2019 and BDD100K further validate the cross-domain generalization capability of the proposed framework.

Abstract: Real-time quantitative precipitation estimation (QPE) from weather radar is essential for hydrological forecasting, flash flood warning systems, and water resource management. Despite significant advances in radar technology and signal processing, operational QPE systems face persistent challenges including non-meteorological clutter contamination, signal attenuation, vertical profile biases, and systematic errors that require integration with ground-based rain gauge networks. This review synthesizes recent developments in open-source frameworks for radar QPE, spanning the complete processing chain from raw signal correction to operative hydrological validation. We examine state-of-the-art methods for clutter removal (polarimetric fuzzy logic, CLEAN-AP, neural network quality control), C-band attenuation correction (self-consistent and KDP-based approaches), and vertical profile of reflectivity (VPR) correction for warm-rain events. We compare gauge-radar merging techniques including mean field bias adjustment, spatially variable corrections, Kriging with External Drift (KED), and Conditional Merging, with emphasis on real-time applicability and look-back window strategies. The review identifies key open-source Python libraries (wradlib, Py-ART, pySTEPS, radproc, weatherDataHarmonizer) and documents operational latency constraints for flash flood warning systems. A critical research gap is identified: current open-source solutions lack documented workflows for integrating delayed 24-hour manual gauge readings into real-time QPE streams while maintaining low latency. This review provides researchers and practitioners with a comprehensive roadmap for developing robust, open-source, real-time radar QPE systems suitable for operational hydrological applications.

Abstract: This study introduces the Children's Empathy for Older Adults (CEOA) eight-item scale, a novel image-based instrument designed to measure young children's views, empathy, and behavioural intentions toward older adults. CEOA was administered as a pre-test and post-test metric, following storytelling sessions, on 232 children aged 5-6 years in the multi-racial and multi-cultural context, Singapore. Findings revealed that children with regular exposure to grandparents demonstrated clearer, more distinct responses across all three domains, indicating a more developed understanding of older adults’ needs. In contrast, children without such exposure showed less differentiation between cognitive, affective, and behavioral components. These results underscore the importance of intergenerational contact in shaping children’s perceptions and empathy for older adults. The CEOA scale is a valuable tool for future research and interventions aimed at fostering positive intergenerational relationships.

Abstract: This paper examines whether the rise of remote work following the COVID-19 pandemic has generated a structural transformation in urban spatial organization across major metropolitan areas in advanced economies. While much of the existing literature treats COVID-19 as a temporary shock, this study argues that it has induced a persistent reconfiguration of cities toward more polycentric and decentralized spatial structures.Using a multi-source dataset combining Google mobility reports, NASA/VIIRS night-time light satellite data, OECD and national labor force surveys, and urban economic indicators, the study constructs a novel Urban Polycentricity Index (UPI) to measure spatial dispersion of economic activity. The empirical analysis covers New York, London, Paris, Berlin, and Munich over the period 2019–2025.The methodology integrates structural break tests, difference-in-differences estimation, and spatial equilibrium modeling to identify both the timing and magnitude of post-COVID spatial shifts. Results indicate a significant structural break around 2020–2021, followed by a sustained increase in remote work adoption and urban polycentricity. Satellite and mobility data confirm a systematic redistribution of economic activity from central business districts toward suburban and peripheral zones.Findings show that remote work is a statistically significant driver of urban decentralization, associated with flatter density gradients, reduced commuting intensity, and higher polycentricity. Counterfactual simulations further confirm that, without remote work expansion, cities would have remained substantially more monocentric. Overall, the study demonstrates that COVID-19 has permanently altered urban spatial equilibrium, positioning remote work as a key structural force reshaping metropolitan form.

Abstract: Integrated care is increasingly shaped by digital infrastructures, data governance, and AI-enabled analytics, yet the relevant literature remains fragmented across health-services research, digital health, and machine learning. This article presents a conceptual review informed by structured scoping searches across PubMed, Scopus, Semantic Scholar, Crossref, and selected policy sources covering January 2001–March 2026. The search component was used to map the field and identify representative frameworks, implementations, and technical advances rather than to estimate pooled effects. We synthesise the literature across four domains: conceptual foundations of integrated care, AI and multimodal analytics, implementation barriers, and digital-governance requirements. On that basis, we propose a five-level taxonomy ranging from disease-specific programmes to learning integrated care models and argue that most current deployments remain concentrated at digitally integrated but only weakly adaptive Type IV configurations. Across the literature, three recurrent constraints limit progression towards Type V learning systems: temporal blind spots, maintenance debt, and governance misalignment. Overall, the review positions AI-enabled integrated care less as a finished model than as an emerging design space requiring longitudinal data assets, stewarded model lifecycles, and accountable governance to support clinically useful, equitable, and trustworthy learning systems.

Abstract: The global drive towards sustainability and energy conservation has accelerated the development of intelligent buildings utilizing building management system (BMS). Occupants have profound impacts on building environment. Incorporating occupant-related factors into the environmental control process is essential for optimizing the efficiency of BMS, which thus give rise to the concept of occupant-centric control (OCC). Conventional methods rely on simplified models and fixed schedules that fail to satisfy environment control and occupant requirements, while constructing credible models places strict requirements on the dataset. In this paper, we propose a Model-Aware Predictive Control framework named MAPC, which can construct credible models with limited data and provide room-level control strategies allowing for occupant comfort and energy efficiency. Its technological innovations are twofold. On the one hand, we design a model construction and fine-tuning method combining data-driven subspace projection approach with physical priors, which can construct credible thermal dynamic models with limited data. On the other hand, to balance the potential conflicts between enhancing occupant comfort and saving energy, we present a hierarchical decision-making mechanism, which enables room-level global optimal control considering dynamic occupant comfort requirements and energy usage. Experimental results obtained on a typical duplex apartment dataset demonstrate that MAPC is able to provide room-level control strategies based on dynamic occupant requirements and user preferences, achieving improved occupant comfort and energy efficiency. The ablation experiments also demonstrated the superiority of MAPC in constructing reliable models on limited datasets.

Abstract: As the global average temperature increases, heat stress (HS) caused by high temperatures has become a key constraint to the development of the poultry industry. As the primary metabolic organ, HS can induce liver injury in chickens, thereby compromising food safety. However, the precise mechanisms underlying HS -induced liver injury remain to be elucidated. The objective of this study is to explore the impact of HS on liver damage, oxidative stress, the Keap1-Nrf2 pathway, ferroptosis and cuproptosis in chickens. A total of 70 chickens were selected for this experiment and divided into a CON group and HS group: the CON group was reared in a normal-temperature environment (24 ± 1 ℃), whilst the HS group was reared in a high-temperature environment (33 ± 1 ℃). The findings of the study suggested that HS has the potential to induce liver dysfunction, oxidative stress, and disruption of the Keap1-Nrf2 pathway. HS has been demonstrated to induce Fe2+ accumulation in chicken livers, inhibit the expression of FTH1, FSP1, SLC7A11 and Gpx4, and simultaneously upregulate the expression of CD71, PTGS2 and ACSL4, thereby promoting ferroptosis. Furthermore, Cu2+ accumulation in the liver upregulates HSP70, DLAT and Lip-DLAT levels and downregulates the expression of ATP7B, PDH1A, PDHB, PDK4, DLST and FDX1, thereby inducing cuproptosis. Subsequent correlation analysis revealed that HS can induce ferroptosis and cuproptosis via the HO-1/FDX1/Gpx4 pathway. This finding provides new insights into the mechanisms underlying HS-induced liver injury.

Abstract: Circadian rhythms are fundamental regulators of bone remodeling, orchestrating the co-ordinated actions of osteoblasts, osteocytes, and osteoclasts. Recent studies have high-lighted how core clock genes, such as Bmal1, Clock, Per1/2, and Cry1/2, exhibit rhythmic expression in bone tissue and modulate key markers of bone formation and resorption. Disruptions in circadian regulations, whether caused by environmental factors or genetic alterations, have been linked to osteoporosis, impaired fracture healing, and increased risk of bone fragility. This review provides a comprehensive evaluation of current experi-mental models used to study circadian regulation in skeletal biology, including in vivo, ex vivo, and in vitro approaches. We summarize their respective advantages and limitations and outline the molecular and cellular markers employed to assess circadian function in bone cells. We also discuss the emerging co-culture models and human-relevant plat-forms, for their potential to bridge the gaps between mechanistic research and transla-tional applications. By comparing model characteristics and highlighting integrated re-search strategies, this review aims to advance circadian bone research and inform future investigations into potential temporal aspects of skeletal health.

Abstract: This paper consolidates, corrects, and extends a research programme on the shifted-prime problem $p = q + r - 1$ with $p, q, r$ prime and its connections to the binary Goldbach conjecture and the non-trivial zeros of the Riemann zeta function $\zeta(s)$. New material over Version 6. The principal addition is a rigorous three-level treatment of the restricted Goldbach sum \[ P_{R_{3,4}}(N) = \sum_{\substack{p+q=N,\\ p\equiv 3\ (\mathrm{mod}\,4)}} (\log p)(\log q). \] At Level 1 [PROVED] (unconditional), the ``almost-all'' theorem of Montgomery--Vaughan type shows that the exceptional set of even integers $N\leq X$ for which $|R_{3,4}(N) - \tfrac{1}{2}C_2 S(N)N|$ exceeds $CN/(\log N)^3$ has measure $O_A\bigl(X/(\log X)^A\bigr)$ for every $A>0$. At Level 2 [PROVED] (unconditional), a transfer inequality bounds $|R_{a,q}(N)-\phi(q)^{-1}R(N)|$ in terms of twisted sums $S_\chi(N)$ with mean-square control. At Level 3 [COND. PROVED, GRH], for all sufficiently large even $N$ one has $R_{3,4}(N)=\tfrac{1}{2}C_2 S(N)N + O(N^{1/2+eps})$. Anderson's original claim of an explicit unconditional constant $K\leq 28.65$ for all $N$ is identified as relying on the Hardy--Littlewood binary asymptotic for each individual $N$, which is itself a conjecture; the claim is accordingly downgraded and the gap stated precisely. Retained from Version 6. Five analytical gaps (A--E) in the Goldbach--Riemann bridge for $\Psi^*(x)$ are fully closed unconditionally (Gaps D1, D2, D3, E) or under GRH (Gap C). The corrected spectral-detection results stand: $\lambda_1/\lambda_2 = 182.63$ ($n=892\,206$); 129/200 Riemann zeros detected at $p<0.01$ ($n=1\,310\,763$); Mellin--Lomb--Scargle concordance 29/30 versus 0/30; 9/10 direct Pearson correlations significant; heteroscedasticity of $eps(p)$ formally confirmed ($p=4.7\times 10^{-14}$). Principal corrections retained from Version 6. The $k=3$ existence problem is equivalent to binary Goldbach (open). The permutation-test bug in scripts~6.py--8.py is corrected ($199/200\to 129/200$). The formula for $S_\infty^{(k)}$ is corrected for $k\geq 3$. None of these results constitutes a proof of the Riemann Hypothesis. All claims carry explicit epistemic labels.

Abstract: Recently, a new generalization of hyperbolic functions called para-hyperbolic functions has been introduced. However, properties of the para-hyperbolic functions have not been investigated yet. In this paper, we derive the correct explicit formulas of the para-hyperbolic sine and cosine, study elementary properties of these functions, and explore which of the inequalities that hold for trigonometric and hyperbolic functions find their counterparts for para-hyperbolic functions. Namely, we prove a Wilker type inequality, Cusa-Huygens and Lazarević type inequality, Wu-Debnah modification of Wilker type inequality and Shafer type inequality for para-hyperbolic functions.

Abstract: The Collatz (3x + 1) conjecture remains one of the most challenging open problems in number theory, largely due to the unpredictable, pseudo-random fluctuations of its discrete integer o rbits. This paper introduces an interdisciplinary approach by translating discrete arithmetic rules into a continuous dynamical sandbox. Specifically, we construct a symbolic analogy between the 3x + 1 map and the Logistic map f (x) = 1 − µx² locked at the superstable period-3 window (µ ≈ 1.7549). By building a customized threshold partition anchored at the unstable fixed point, the continuous system naturally enforces a “forbidden word 11” grammar, mirroring the arithmetic constraint that an odd operation (T(n) = 3n + 1) must produce an even number. Through the eigenspectrum of the Perron-Frobenius transfer operator, we demonstrate a 2:1 ergodic measure ratio for contraction (even) and expansion (odd) states—a direct geometric consequence of the period-3 attractor structure. We validate the ro-bustness of the spectral quantities through convergence studies across multiple discretization schemes. Null-model controls show that the sandbox captures aspects of the global stopping-time distribution that a generic forbidden-11 Markov chain does not, while run-length analysis reveals that local arith-metic statistics (ν2-valuations) are better reproduced by the simpler null model. This mixed result delineates the sandbox as a partial surrogate: useful for global transient statistics, but not a replacement for the actual arithmetic dynamics. This study offers a heuristic framework positioning coarse-grained transient dynamics as a null-model approach for Collatz statistics, with explicitly characterized failure modes.

Abstract: The bioactive plant peptides represent a significant and yet largely underexploited resource with huge potential not only for basic plant science but also for various biotechnological applications, including pharmaceutical and agrochemical development. bioactive plant peptides represent a significant and yet largely underexploited resource with huge potential not only for basic plant science but also for various biotechnological applications, including pharmaceutical and agrochemical development. This fastexpanding research area of plant peptidomics demands the creation and continuous updating of dedicated databases that facilitate data integration of heterogeneous nature and enable efficient knowledge discovery. The most representative databases, like PlantPepDB and PhytAMP, but also recent multi-purpose databases like MFPPDB, are reviewed here in terms of data sources used-literature and public repositories-manual curation extent, functional classification-such as therapeutic, defense, and inhibitory-and the availability of relevant metadata on physicochemical properties and structure. While databases focused on specific bioactivities of plant peptides offer high-quality, focused data, broader repositories are crucial for discovering multifunctional peptides and structure-activity relationships. The refinement and integration of these databases, alongside advanced bioinformatics tools, remain essential for overcoming these hurdles. These resources stand to facilitate innovation in ways that will continue to illuminate insights into the molecular function of plants and allow the successful harnessing of plant peptides toward human health improvements and sustainable agriculture. This review briefly introduces the progress of plant peptide research, presents an overview of plant peptide studies, and provides a comprehensive analysis of existing plant peptide databases, evaluating their scope, content, and utility. We anticipate that this work will bridge the gap between peptide discovery and the development of nextgeneration plant peptide databases.

Abstract: We introduce a new notion, that of a contextuality profile of a system of random variables. Rather than characterizing a system's contextuality by a single number, its overall degree of contextuality, we show how it can be characterized by a curve relating degree of contextuality to level at which the system is considered, \( \begin{array}{c|c|c|c|c|c|c|c} \textnormal{level} & 1 & \cdots & n-1 & n>1 & n+1 & \cdots & N\\ \hline \textnormal{degree} & 0 & \cdots & 0 & d_{n}>0 & d_{n+1}\geq d_{n} & \cdots & d_{N}\geq d_{N-1} \end{array} \), where N is the maximum number of variables per system's context. A system is represented at level n if one only considers the joint distributions with \( k\leq n \) variables, ignoring higher-order joint distributions. We show that the level-wise contextuality analysis can be used in conjunction with any well-constructed measure of contextuality. We present a method of concatenated systems to explore contextuality profiles systematically, and we apply it to the contextuality profiles for three major measures of contextuality proposed in the literature.

Abstract: Companion dogs are increasingly recognized as translational models for studying human physiology and disease. Unlike conventional or genetically engineered laboratory models, dogs are outbred, immunocompetent animals that spontaneously develop complex diseases whose pathogenesis and environmental exposures commonly overlap with those of humans. These distinctive features create opportunities to study mechanisms of disease, progression, and therapeutic responses under conditions that more closely resemble clinical reality. This review highlights evidence for the translational relevance of canine models across multiple therapeutic areas. We further discuss how advances in genomics, transcriptomics, spatial biology, in vitro, and in silico model systems are expanding the translational utility of canine models for applications in human medicine. Although important species differences must be carefully weighed, dogs represent a uniquely valuable comparative model for elucidating disease mechanisms, informing drug development, and accelerating the translation of scientific discoveries to human medicine.

Abstract: Oleoresins are complex natural lipophilic matrices traditionally analyzed using chromatographic techniques that require extensive sample preparation, derivatization, and authentic standards. Amazonian oleoresins from Copaifera and Eperua species (Fabaceae) represent valuable bioresources with recognized pharmacological potential, largely attributed to diterpenoids such as copalic and hardwickiic acids, as well as bioactive sesquiterpenes, including the cannabinoid b-caryophyllene. In this study, we present a proof-of-concept application of Direct Analysis in Real Time coupled with High-Resolution Mass Spectrometry (DART-HRMS) as a rapid, direct, and environmentally friendly approach for chemical fingerprinting and semi-targeted screening of the two most important amazonian oleoresins from these two genera: Eperua oleifera and Copaifera multijuga. Analyses were performed using a Q Exactive Orbitrap coupled to a DART ion source under after conditions optimization. Hardwickiic acid was used as a model compound for method optimization, with optimal performance achieved at 200 °C and 100 V, yielding stable signal intensities (CV &lt; 10%) and high mass accuracy (&lt; 1 ppm). The method enabled reproducible detection of diterpenic acids in both oleoresins, allowing differentiation of their chemical profiles and assessment of short-term stability under ambient conditions. In addition to diterpenes, free fatty acids were also detected, expanding the compositional characterization of these matrices. Compound annotation was performed based on accurate mass measurements and literature comparison, corresponding to Level 5 confidence according to established metabolomics criteria. Although the absence of chromatographic separation limits isomer discrimination and absolute quantification, DART-HRMS provides a rapid and solvent-free strategy for chemical fingerprinting and preliminary characterization of oleoresins. This approach aligns with Green Chemistry principles and shows strong potential as a screening and triage tool for quality control, chemotaxonomic studies, and sustainable valorization of Amazonian natural products.

Abstract: Ear biometrics has emerged as a promising alternative in biometric recognition systems, offering robustness in unconstrained environments where traditional modalities such as face recognition may fail on its own, but can be enhanced by ear. Ear segmentation, in particular, plays a crucial role in downstream recognition by isolating discriminative ear regions and reducing background interference. However, existing approaches to ear detection and segmentation are commonly susceptible to severe occlusions, ear accessories, and variable illumination, and their performance deteriorates on images captured in the wild. To address these limitations, we introduce a tailored ear-segmentation architecture based on a U-Net with a ResNet-50 encoder. Trained and validated on the Annotated Web Ears (AWE) dataset, our method achieves a mean Intersection over Union (IoU) of 77.1% and a pixel-wise accuracy of 99.7%, outperforming the Convolutional Encoder--Decoder (CED) baseline. We further evaluate on the EarSegDB-25 dataset, where our approach attains a test-set IoU of 94.76%, significantly surpassing previous ear segmentation methods based on the original U-Net architecture. High pixel-wise accuracy across methods is largely attributable to background dominance; in contrast, the improved IoU achieved by our approach more accurately reflects gains in ear region segmentation performance. Leveraging a ResNet-50 encoder, our model demonstrates robust performance under occlusion and illumination challenges, achieving state-of-the-art results on AWE and EarSegDB-25 and showing strong potential for biometric applications in unconstrained environments.