Accurate food portion size estimation is a critical challenge in nutrition analysis and dietary assessment. Recent 3D reconstruction methods primarily focus on surface geometry, often neglecting the volumetric accuracy necessary for precise portion size estimation. In this work, we present three methods, NS-DRS, HR-CMS, and GS-GP, for real-scale 3D food reconstruction and volume estimation from limited 2D inputs. All three methods follow a shared three-stage pipeline consisting of 3D reconstruction, scale estimation using physical references such as visible checkerboards, and mesh refinement. While this high-level structure is consistent, the methods differ in their reconstruction backbones and scale-recovery mechanisms, and are designed for different input conditions, including both single-view and multi-view images. We evaluate each method on a diverse set of food items exhibiting variations in texture, shape, and camera pose, and assess their performance in terms of volume estimation and geometric accuracy. The results reveal complementary strengths among the pipelines, with NS-DRS performing better in volume estimation and GS-GP being more effective in 3D reconstruction. Across all settings, the three methods achieve 18-23% lower volume estimation error compared to the current state-of-the-art model. These findings demonstrate the effectiveness of physically informed and explainable reconstruction pipelines for accurate portion estimation, and support their potential use as scalable tools for dietary monitoring and clinical nutrition analysis.

The population of Wallago attu (Bloch & Schneider, 1801) has declined worldwide including in the rivers of India and has been declared ‘Vulnerable’ by the International Union for Conservation of Nature. The restoration and conservation of fishes is heavily dependent on their population status in their natural habitats. Therefore, otolith shape was used to study the stock structure of W. attu from rivers Ganga, Gomti and Yamuna in Northern India, river Hooghly in Eastern and Pampa in Southern part of India. Shape indices of the asteriscus otoliths and elliptical Fourier analysis of the contour shapes was investigated using univariate and multivariate statistical techniques. The otolith descriptors (otolith length, breadth, area and perimeter) and four shape indices (length-breath ratio, form factor, circularity and ellipticity) of the otoliths were different (ANOVA, p<0.05) between the fish of five sampled rivers. Principal component analysis of normalized elliptical Fourier descriptors (NEFDs) of the contour shape accounted for 89.55% of the total variance. Discriminant function analysis on the basis of NEFDs resulted in 100% classification rate with no intermixing between the fish of different rivers and depicted the presence of different stocks of W. attu in a scatter plot of DF-I against DF-II. These variations in the otolith shape in the current study might be due to the restricted movement of fish because of geographical isolation and different environmental conditions in different rivers due to changes in habitat, or both. The current study would be helpful in developing effective strategies for stock management of this species.

Leaf shape displays remarkable diversity, with its evolution hypothesized to reflect adaptive ecophysiological functions. Theoretical models propose that variation in leaf shape-particularly through modifications in effective leaf width (we)-primarily influences thermoregulation and hydraulic efficiency. However, comprehensive empirical tests of these hypotheses are lacking. Oxytropis diversifolia E. Peter (Fabaceae) has natural variation in leaf shape (1 leaflet, 1-3 leaflets, and 3 leaflets) and exhibits clinal variation, making it an ideal candidate to test those functional relationships. Here, we quantified leaf morphometrics across populations, logged in situ leaf temperature and gas exchange, and examined leaf anatomy associated with water balance. We confirmed that the production of more leaflets did reduce we. While leaves with reduced we could stay cooler during the day, the extent of leaf-to-air temperature difference was typically small (often within 1°C), suggesting a limited biological impact. Crucially, we identified a key anatomical trade-off in water relations: reduced we yielded beneficial lower chlorenchyma-to-midrib ratios and higher vein density, but at the cost of smaller vascular dimensions. This trade-off likely underpins the observed, context-dependent superior gas exchange of the intermediate phenotype. We propose that the functional significance of leaf shape lies in water relations over thermoregulation, with balancing selection on the anatomical trade-off providing a plausible mechanism maintaining the polymorphism.

Designing a quad mesh that meets aesthetic, anatomical, and numerical requirements often requires meticulous manual effort in conventional methods, making quadrilateral remeshing an “art of design”. Neural networks hold significant promise for automating this process. However, current approaches that directly predict cross fields cannot properly handle the discontinuous behavior of smooth cross fields: minor shape variations can lead to substantial changes in the cross field, even when singularities remain largely unchanged. Therefore, such methods often result in non-smooth outputs when combining multiple singularity instances. To avoid such discontinuity, we propose to learn the sparse singularities, including their locations and indices, then let the non-neural conventional method to smoothly connect them. The imbalanced ratio of singular and regular vertices poses a significant challenge for learning. We convert them into a geodesic distance field and an over-sampled index field to address it. This carefully designed two-stage strategy satisfies several key requirements, such as coordinate invariance and tessellation insensitivity, while enabling the generation of smooth cross fields with varying topologies. By shifting the focus from directly learning the cross field to learning singularities, we also simplify the dataset preparation process by requiring only sparse annotations.

Accurate and efficient surface defect detection is crucial for ensuring the structural integrity and quality of industrial steel products. However, existing detectors still struggle with sub-pixel, low-contrast defects and extreme variations in defect scale and shape. To address these challenges, we propose TriA-Mamba-You Only Look Once (YOLO), a triple-awareness enhancement of the Mamba-YOLO framework for steel defect detection. First, the input samples are processed using a discrete wavelet transform, which reduces background noise and enhances the representation of defect features. Next, the backbone is enhanced with a micro-feature-aware enhancement block, which comprises a Smooth-Mix Focus Stem (FS) block to minimize aliasing artifacts while preserving critical defect features, combined with a Selective Fusion Attention (SF-Att) block for refinement by selectively emphasizing discriminative micro-scale patterns. Within the neck, a context-aware feature enhancement module is implemented to tackle extreme scale and shape variance. It utilizes a receptive-field lightweight spatio-channel feature refinement (RLSC-FR) block to optimize local structure and receptive field characteristics, and an Attention-Augmented Visual State-Space (AA-VSS) block to dynamically adapt to defect scale and morphological variations. Finally, a path-importance-aware fusion module, using Norm-Weighted concatenation (NW-Concat), is developed for multi-path feature fusion, dynamically adjusting contribution weights to ensure balanced aggregation. Experimental results show that TriA-Mamba-YOLO achieves a mean average precision (mAP) of 81.3%, surpassing the original Mamba YOLO by 5.2% on the Northeast University-Defect Detection (NEU-DET) dataset, and achieves a balance between accuracy, computational efficiency, and robustness. These results highlight the effectiveness of the proposed approach for industrial surface defect detection applications.

Background: Coronavirus disease 19 (COVID-19) 1st emerged in Wuhan province of China, caused by the SARS-Cov-2 virus in December 2019. COVID-19 has systemic manifestations, including hematologic manifestations. Complete blood count (CBC) has been recommended as a valuable tool for patient monitoring.Methodology: Our study was an ambispective study on COVID-19-positive patients whose complete blood count samples and peripheral smears. We analysed 700 patients with COVID-19 and recruited perfectly matched controls (COVID Negative) of similar age and gender. Result: The study included 700 cases (125 cases CBC with peripheral smears and the remaining 575 with only CBC) with an additional 60 COVID PCR negative and 60 COVID-recovered patients. COVID-19-positive patients had a mean age of 49.5 years with an M: F ratio. ESR mean value was 36mm in 1st hr. TLC count of 10.06 *103/µL whereas the COIVD recovered, and COVID-negative patients had TLC in the mean of 7.32*103/µL & 8.03*103/µL and was significant (p-value 0.000). A comparison of leucocyte counts among the three groups showed that the neutrophil, lymphocyte, monocyte, eosinophil count and their absolute values being statistically significant. Platelet count, MPV, and PDW were also significant among the three groups of patients. Discussion & Conclusion: Peripheral smear evaluation of COVID-19 patients could provide an insight into the severity of disease and multi-organ involvement. Morphological analysis by a trained pathologist cannot be entirely replaced by artificial intelligence. When looked at by a trained pathologist, subtle shape, size, and morphology variations can help adequately manage patients.

Leaf shape in plants plays a vital role in water use, tolerance to abiotic stresses, and formation of canopy structures. Muscadine grapes ( Vitis rotundifolia ) are native to the southeastern United States and typically have unlobed leaves. A previous breeding effort successfully introduced the lobed-leaf trait of Vitis vinifera into the muscadine cultivar Southern Home via a series of interspecific hybridizations. A linkage-based high-density genetic map was developed using a biparental mapping population from a cross between the muscadine selection Ga. 8-1-313 (female, unlobed leaves) and ‘Southern Home’ (hermaphroditic, lobed leaves) to map quantitative trait loci (QTLs) associated with leaf lobing. The mapping population consisted of 130 F 1 progeny that were genotyped via a genotyping by sequencing method (GBSpoly). Single nucleotide polymorphism (SNP)-based genetic maps were constructed and verified by mapping the grape sex locus. Mature leaf samples from both parents and progeny were collected, digitally scanned, and 17-point landmarks were placed in well-defined biologically homologous points of each leaf image. The x, y coordinates of these landmarks were then used to analyze leaf shape using generalized Procrustes analysis (GPA). Significant variations in leaf shape were identified via principal component analysis (PCA). The principal component (PC) scores were used as phenotypes to perform QTL analysis, and a major QTL determining leaf lobing was identified on chromosome 1 in ‘Southern Home’, which explained 75.1% to 76.6% of total phenotypic variance. Further study of the identified QTL region can be carried out to identify candidate genes involved in leaf lobing in grapes.

Projection-based tokenization with Pseudo feature learning for ovarian lesion segementation in ultrasound images.

Diffusion probabilistic models (DPMs) have recently demonstrated promising performance in medical image segmentation. However, traditional DPM has difficulty handling the irregular structure of images and the inherent similarity between lesions and surrounding tissues. To overcome these challenges, we propose an innovative architecture, the multiscale Mamba DPM (MsM-DPM), designed to enhance medical image segmentation. Specifically, MsM-DPM introduces a multiscale attention fusion module (MSAFM) in a multiscale denoising UNet (Ms-DU) to capture lesion deformations from multilevel features, thereby enhancing the model's robustness to shape and scale variations. Furthermore, in the segmentation network, a multilayer axial feature module (MLAFM) is used to adaptively aggregate the global context features from the Mamba encoder to enhance the expression of features in the spatial dimension by capturing axial multiscale features. The multilevel global context (MLGC) module is then used to reconstruct skip connections using graph convolutional network inference, and the enhanced features are assigned to each layer in the decoder to capture the contextual relationship of features. Finally, the feature fusion module (FFM) integrates deep features with upsampled features in the decoder, enhancing the network's ability to capture lesion boundary details. Our MsM-DPM effectively encodes the semantic difference between lesions and background to improve the representation of their internal features. Extensive experiments on six datasets, LUNA16, ATM22, COVID-19, Self-collected datasets, Pancreas, and BT-MSD, show that the proposed MsM-DPM outperforms existing segmentation methods. Our code is publicly available at https://github.com/suhuaqiang/deep-learning.

Abstract Understanding variations in the dust extinction curve is imperative for using dust as a tracer of local structure in the interstellar medium, understanding dust chemistry, and performing extinction corrections. However, the extinction curve is complicated and exhibits features across a wide range of wavelength scales, from narrow atomic lines and diffuse interstellar bands (DIBs), to intermediate-scale and very broad structures (ISSs and VBSs), and the overall slope of the optical extinction curve, parameterized by R ( V ). Robust, population-level studies of variations in these features are only now possible with large, all-sky, spectroscopic surveys. However, these features are often studied independently because they require drastically different spectral resolution. In this work, we couple features with disparate wavelength scales by cross-matching precision catalogs of DIB measurements from APOGEE (7 DIBs) and Gaia RVS (1 DIB) with low-resolution extinction-curve measurements from Gaia XP. Using this combination, we demonstrate meaningful correlations between the strengths of extinction-curve features across all wavelength scales. We present a model that partially explains the excess scatter in DIB strength versus extinction and we show variation in line shapes of two DIBs as a function of R ( V ). We find that most DIBs studied in this work increase in strength with increasing R ( V ) and/or increasing ISS strength, though we found one DIB that anomalously decreases in strength with increasing R ( V ). Using the behavior of the ensemble of APOGEE DIBs, we present this as the first evidence of systematic chemical composition variation accompanying R ( V ) variation.

ABSTRACT Skyrmion systems have been regarded as potential candidates for versatile energy‐efficient information processing due to the intrinsic topological properties. In emerging skyrmion‐based reservoir computing concepts, history‐dependent spin state evolution constitutes a key physical ingredient, highlighting the importance of controllable collective dynamics in response to external stimuli. Antiskyrmions, as antiparticles of skyrmions, are expected to offer additional configurational degrees of freedom and enhanced thermal stability. However, experimental visualization of their field‐dependent collective evolution remains scarce. Here we report a continuous, field‐history‐dependent antiskyrmion transition from a triangular to square lattice, accompanied by a sequence of intricate intermediate states in Mn 1.4 PtSn chiral magnet. Coordinated variations in antiskyrmion shape, size, and position are directly demonstrated, thereby offering experimentally multiple accessible degrees of freedom under controlled magnetic‐field inputs. Systematic micromagnetic simulations reveal that the competition among Dzyaloshinskii–Moriya, dipolar, and Zeeman interactions governs the sequential reconfiguration of local spin textures underlying the observed lattice evolution. Our results provide a controllable and history‐dependent antiskyrmion lattice platform with rich intermediate configuration states to explore multi‐level information encoding and reservoir‐computing applications.

ABSTRACT Menopause indicates the end of the female reproductive lifespan. The hormonal shifts prior to this event result in a frequently symptomatic period known as perimenopause, which has been linked to unfavorable changes in body composition. This study explored women’s experiences of perimenopause, including perception of body shape and weight. Perimenopausal women living in Ireland were recruited via convenience sampling to participate in semi-structured virtual interviews with a focus on experiences of weight and body shape. Interviews were recorded, transcribed and subjected to thematic analysis. Seventeen participants with an average age of 50.6 ± 2.8 years participated. Six main themes were produced, describing the far-reaching impact of menopause and its symptoms. The women expressed a differing experience of weight and body shape during perimenopause, compared to previous experiences. They were attempting to make sense of body shape and weight variations, with body image often negatively affected. Women aspired to make lifestyle changes, driven by a desire to manage weight, reduce symptoms, or improve present/future health. Challenges specific to this life stage, including symptoms or additional life factors, act as barriers to engaging with healthy lifestyle strategies. Women are seeking guidance to navigate perimenopause and any weight and body shape changes indicating that targeted lifestyle interventions are required, with a role for supportive weight management strategies.

Brain tumor classification using magnetic resonance imaging (MRI) presents challenges due to variations in tumor size, shape, and texture. Although traditional image preprocessing methods are commonly employed to improve input quality, their impact on optimizer behavior and CNN performance has yet to be thoroughly investigated. This research examines the effect of preprocessing on convergence, generalization, and classification accuracy across various optimizers. We utilize a publicly available Kaggle dataset to create two preprocessing pipelines: a baseline pipeline that only resizes images and a traditional pipeline that converts images to grayscale, blurs them, and applies morphological filtering. We then test how these pipelines affect three optimizers: Adam, Root Mean Square Propagation (RMSProp), and Stochastic Gradient Descent (SGD). To separate protocol variables, a fixed CNN architecture is used throughout. Performance is assessed using accuracy, precision, recall, and F1-score, validated through five-fold cross-validation. Results show that baseline preprocessing consistently yields higher accuracy and more stable convergence across all optimizers, with RMSProp and SGD achieving the highest mean accuracy of 99.53% under five-fold cross-validation. The findings address the understudied effect of preprocessing on optimizer performance, emphasizing the need for preprocessing-aware training strategies to improve robustness and interpretability in medical image analysis.

Abstract The effect of plasma shaping on the $n=4$ resonant magnetic perturbation (RMP) response in EAST is investigated using the resistive magnetohydrodynamic code MARS-F (Liu \textit{et al} 2010 \textit{Phys. Plasmas} \textbf{17} 122502) under both upper and lower single null (USN and LSN) configurations, with the aim of guiding experimental shaping strategies for enhanced control of edge-localized mode (ELM). Upper and lower triangularity ($\delta_U$,$\delta_L$) or elongation ($\kappa_U$,$\kappa_L$) are varied independently via a coordinate transformation that closely mimics the experimental shaping, while key equilibrium parameters are held fixed. The simulations reveal that increasing the triangularity or elongation on the X-point side (upper in USN, lower in LSN) consistently enhances the edge plasma response, as characterized by increased resonant radial field ($b_{res,outer}$), stronger edge peeling response ($A_{peel}$), and intensified surface displacement near the X-point ($\xi_X$), all of which are favorable for ELM suppression. Conversely, increasing shaping parameters on the non-X-point side (lower in USN, upper in LSN) weaken these responses. The optimal RMP coil phasing for maximizing plasma response exhibits an approximately linear dependence on the shaping parameters, with greater sensitivity to variations on the X-point side. Notably, when plasma shaping on X-point side is weak, the optimal phasing derived from $\xi_X$ deviates from those based on $b_{res,outer}$ and $A_{peel}$, suggesting increased complexity in determining the optimal phasing for experiment; in contrast, such divergence does not occur when shaping is varied on the non-X-point side. The shaping variation approach proposed in this work offers a theoretical basis for optimizing plasma geometry and RMP configuration to improve ELM control efficiency in future fusion reactors.

Abstract Animals can dramatically alter ecosystem structure and function through the cycling and transport of nutrients in their waste. While birds are particularly capable of influencing nutrient cycles due to their high mobility, abundance, metabolism and functional diversity, the factors shaping variation in nutrient release among birds remain poorly understood. We examined how trophic position, phenology and body size shape the release of carbon (C), nitrogen (N) and phosphorus (P) in the waste of a diverse assemblage of songbirds. We analysed waste samples (reflecting excretion and egestion) from 151 individual songbirds (31 species) at a migratory stopover site during one spring and one fall season in upstate New York, USA. Trophic position, represented by δ 15 N in waste, correlated positively with %N and negatively with C:P. Waste stoichiometry also differed between seasons with higher %N and lower C:N in spring compared to fall as birds shifted their diets from N‐rich insects in spring to C‐rich fruits and seeds in fall. Birds in the fall exhibited increased fattiness, and body size had the strongest influence on C, N and C:N release during this season, suggesting that fat accumulation for energy storage may shape nutrient excretion prior to migration. However, the effects of body mass and its interactions with season and trophic position were complex. Ultimately, trophic position, phenology and body mass together helped explain variation in nutrient release among songbirds. Future research may leverage this information to further explore how migratory songbirds influence stopover habitats and other ecosystems via contributions to nutrient cycling. Read the free Plain Language Summary for this article on the Journal blog.

Abstract We build a multilayer magnetohydrodynamic shallow-water model to study the thickness and shape of the solar tachocline. This allows us to include characteristics of both the overshoot and the radiative parts of the tachocline. The equations derived include equilibrium in latitude among Coriolis, pressure gradient, and magnetic curvature stresses for each layer, and magnetohydrostatic equilibrium in the radial direction. In each layer, the total mass is conserved; mass is redistributed for different amplitudes and latitude positions of toroidal bands, thus producing variations in tachocline shape and thickness with solar cycle phases. While we solve here for equilibrium of two layers, the equations can be readily generalized for additional layers. In pure hydrodynamic tachocline with no differential rotation, thickness and shape are independent of latitude. With differential rotation and/or magnetic fields, the tachocline is, in general, oblate in equatorial regions but prolate in polar latitudes. A local bump occurs at the poleward side of tachocline toroidal band. Hence, depending on latitude-location and amplitude of magnetic band as function of solar cycle, the local bump drifts equatorward trailing the magnetic field. Oblateness and prolateness are much larger in the overshoot than in the radiative layer, due to its lower effective gravity. Our results can provide guidance for interpreting helioseismic estimates of variations in tachocline shape and thickness in latitude, including upper limits to banded toroidal field amplitudes.

Everyday behaviour is comprised of myriad components that must be seamlessly coordinated for action to be effective. Individual differences, specifically variation in mental health symptoms, influence how this challenge is navigated, however, their impact in naturalistic settings remains unclear. Adopting a dynamical systems perspective, here we examined whether subclinical variation in symptoms associated with social anxiety disorder (SAD) and autism spectrum disorder (ASD) modulate individual movement dynamics during an everyday activity - walking on a university campus. Participants (n = 93) completed two walking trials, the second of which included an additional distractor task that they were either told to ignore or attend to. Gait dynamics were captured unobtrusively and assessed at both local (i.e., moment-to-moment) and global (i.e., time invariant) levels. The results revealed that subclinical variation symptoms of ASD were associated with less stable local dynamics, independent of task context. Further, exploratory analyses suggested that instructions to ignore the distractor were associated with changes to local dynamics for symptoms of SAD but global dynamics for symptoms of ASD. Taken together, these findings highlight how individual differences in psychological factors can shape the dynamics of everyday behaviour in context-dependent ways.

Differences in social, cultural, and service contexts between urban and rural areas in Indonesia shape variations in how health is socially perceived. This article aims to examine how social perceptions of health are constructed in urban and rural communities and to identify the factors influencing these differences. A qualitative approach employing a structured narrative literature review was applied to twenty scholarly articles published over the past decade, which were analyzed through thematic synthesis. The main findings indicate that urban communities tend to conceptualize health within a rational and biomedical, individual-oriented framework, whereas rural communities interpret health in a more communal and context-sensitive manner, influenced by cultural values, social relations, and local authority. Distinctions are also observed in access to health information, perceptions of disease risk, patterns of preventive behavior, and the strength of social norms. This study concludes that health perception is a socially constructed phenomenon shaped by the interaction of structural conditions, cultural orientations, and social capital. The findings imply that health promotion and intervention strategies should be context-sensitive and tailored to the social characteristics of each setting to achieve sustainable outcomes

Lexical variation plays a pivotal role in linguistics by facilitating the codification of new lexical items, the emergence of New Englishes, and the global recognition of World Englishes. This study presents a comprehensive bibliometric analysis of Scopus-indexed research on lexical variation to identify key trends, thematic focuses, and emerging directions within the field. Using R-based bibliometric tools, the dataset of international publications was systematically analyzed to reveal influential authors, dominant themes, and leading journals. Dialectology emerged as the most extensively studied theme, Dirk Geeraerts as a prolific author, and the International Journal of Corpus Linguistics as a leading outlet. The United States leads in publication output and citation impact, while the Federal University of Bahia is among the most influential affiliated institutions. Co-citation analysis demonstrates the intellectual structure and interdisciplinary connections in the field. The findings highlight growing scholarly interest in lexical variation and underscore the role of sociocultural factors in shaping linguistic identity, innovation, and variation. Future research is encouraged to explore the social dimensions of lexical diversity and examine how technological platforms, including AI- and IoT-driven learning environments, influence language use, emphasizing the need for linguistically inclusive systems.

Abstract Scour holes are common features in river deltas formed by a combination of hydrodynamic and geotechnical conditions. While previous studies have primarily focused on their long‐term evolution toward equilibrium, their response to a single flood wave remains relatively underexplored. This study makes use of a unique, extensive dataset of frequently measured bed levels in the Dutch Rhine Delta to investigate the short‐term behaviour of scour holes. The analyses reveal that scour hole behaviour can be divided into two categories: dynamic and stable . We propose a conceptual model to describe this behaviour. Dynamic scour holes show large fluctuations in area, depth and volume caused by seasonal discharge variations. These temporal fluctuations exceed annual growth rates. By contrast, stable scour holes show virtually no variations in shape and dimensions. They are generally smaller in area but almost twice as deep as dynamic scour holes. Their large depths and steep slopes prevent sediment from being transported up the slopes, showing a stabilisation in depth. Dynamic scour holes are typically located in upstream river reaches, where the sandy beds and engineering works allow for significant erosion and aggradation. Stable scour holes are found in tidal river reaches, where peak discharges are attenuated. Due to the absence of these peak discharges, combined with a strong heterogeneous river bed composed of easily and hardly erodible layers, growth in area is limited.