A mechanistic digital twin of ovarian aging integrating follicular dynamics, mitochondrial decline, and lifestyle perturbations.

Adaptive accelerated failure time modeling with a semiparametric skewed error distribution

Extreme weather events, like drought and heat stress, make it harder to meet water needs in water-insecure settings, particularly vulnerable groups. This study examines how short-term (heat stress) and long-term (drought) water stress affects hydration status across reproductive states (pregnant, lactating, compared to non-pregnant/non-lactating) for Daasanach semi-nomadic pastoralist women in northern Kenya. Drawing on unbalanced panel data, we analyzed 565 observations from 303 women (aged ≥ 16 years) in 2019-2024. Hydration was assessed via urine specific gravity (USG) with dehydration classified as USG > 1.020. Environmental heat stress was measured by ambient temperature and humidity, with sensitivity analyses using wet bulb globe temperature. Mixed effect logistic regression models indicated ambient temperature and humidity were significantly associated with greater odds of dehydration across all women. Holding heat stress constant, lactating but not pregnant women had higher odds of dehydration than non-pregnant/non-lactating women. A significant interaction between heat stress and reproductive status indicated that the probability of dehydration increased fastest for pregnant women as temperatures rose. Holding heat constant, dehydration probability increased during drought years compared to pre- and post-drought and was most pronounced among lactating women. Ambient heat stress increases dehydration risk among Daasanach women with effects compounded in pregnancy, though overall lactation was the period of greatest vulnerability to dehydration. Dehydration probability peaked during the drought illustrating how long-term periods of water scarcity also challenge water needs. Heat stress and droughts exacerbate maternal and infant health risks; thus, targeted hydration and cooling interventions are needed.

• Extension of the ASRF/Vasicek framework to credit portfolios containing two exogenous classes of loans (green and brown), allowing distinct return distributions and correlation structures within a unified asymptotic model. • Incorporation of skew-normal borrower return dynamics, providing a flexible structural mechanism to model empirically documented asymmetries in log-returns while preserving analytical tractability. • Generalization of Vasicek’s convergence result to heterogeneous portfolios with nonuniform exposures, including empirically motivated power-law concentration structures. • Integration of Green Loan Principles (GLP) as the practical basis for loan-type classification, clarifying the financial rationale for distinguishing green and brown exposures. • Scenario-based numerical analysis illustrating loss distributions, sensitivity to key parameters (PDs, correlations, skewness), and the impact of portfolio composition under realistic concentration patterns. • Assessment of robustness of the extended ASRF formula by comparing asymptotic results to finite-portfolio simulations under heterogeneous and skewed structural specifications. • Methodological contribution with regulatory relevance, showing how a green-brown differentiation could be embedded in the ASRF architecture once borrower-level data become available. Inspired by the classical Vasicek approach to credit risk, we propose an extended model for portfolios composed of green and brown loans, widening the Asymptotic Single Risk Factor framework via a two-factor copula structure. Systematic risk is modeled using potentially skewed distributions, allowing for asymmetric creditworthiness effects, while idiosyncratic risk remains Gaussian. Under a non-uniform exposure setting, we establish convergence in quadratic mean of the portfolio loss to a limit reflecting the distinct characteristics of the two loan segments. Numerical results confirm the theoretical findings and illustrate how value-at-risk is affected by portfolio granularity, default probabilities, factor loadings, and skewness. Our model accommodates differential sensitivity to systematic shocks and offers a tractable basis for further developments in credit risk modeling, including granularity adjustments, collateralized default obligations pricing, and empirical analysis of green loan portfolios.

In medicine, multiple continuous outcomes are often repeatedly measured on each subject over time to assess disease severity. Usually, it is of interest to investigate the association between those outcomes, which may be measured at different time points, resulting in unbalanced data. The multivariate linear mixed-effects model (MLMM) is a popular framework for this analysis. It considers the unbalanced nature of the data and accounts for the association of the outcomes via the random effects, often assuming a multivariate normal distribution. However, measuring and understanding the degree of connection between longitudinal outcomes remains challenging. We propose to enhance the MLMM by incorporating various interpretable association structures. Specifically, we consider that multiple longitudinal outcomes are related to the primary outcome through their current value, cumulative effect (total or partial), or both. Our research is motivated by Pompe disease, a rare, inheritable, progressive metabolic myopathy. Clinically, it is important to investigate how patient-reported outcome measures (primary outcomes) are associated with physical outcomes to determine whether improvements in physical outcomes are accompanied by improvements in health-related quality of life and other patient experiences. We found a positive association between them. The proposed models are fitted under the Bayesian framework using Hamiltonian Monte Carlo.

The association between fractional exhaled nitric oxide (FeNO) and airway inflammation is evident. However, the precise relationship of FeNO with pulmonary health and all-cause mortality among participants without airflow limitation remains undisclosed. We investigated the association of FeNO with respiratory symptoms, lung function, and all-cause mortality in this population. Participants included in the 2007-2012 National Health and Nutrition Examination Survey cycles with complete questionnaire information, quality-controlled prebronchodilator spirometry data, acceptable FeNO data, and full follow-up records until December 31, 2019, were included. The skewed distribution of FeNO was addressed by applying natural logarithmic transformation. Multivariable linear regression, logistic regression, and Cox proportional-hazards regression analyses were used to investigate the relationship of FeNO with spirometry, respiratory symptoms, and all-cause mortality. Subgroup analyses were performed based on sex, age, body mass index, smoking status, and blood eosinophil count to validate the robustness of the results. The data of 5,842 eligible participants were analyzed. After adjusting for confounding factors, for each 1unit increment in ln (FeNO), the risk of chronic cough and wheezing decreased by 28% and 22%, respectively. Additionally, forced vital capacity increased by 27.9 mL, and forced expiratory volume in 1 second increased by 27.8 mL. During the average follow-up of 10 years, 255 participants experienced mortality. There was a non-linear relationship between FeNO and all-cause mortality. Specifically, when ln (FeNO) was <2.6 (FeNO < 13.5 ppb), the hazard ratio was 0.45 (95% confidence interval 0.29-0.70; p < 0.001). The subgroup analyses demonstrated consistent results. Elevated FeNO was closely associated with fewer respiratory symptoms and improved lung function in a population without airflow limitation. A non-linear relationship existed between FeNO and all-cause mortality, with mortality initially decreasing as FeNO increased, followed by stabilization.

A robust ATUB-Net for bearing fault diagnosis under unbalanced sample scenarios.

The evaluation of disability grades in traffic accidents is a professional forensic clinical appraisal matter, and its results directly affect the fairness of judicial compensation. In the construction of automated disability grade evaluation models, the imbalanced distribution of disability cases leads to low recognition accuracy for minority categories, becoming a key bottleneck restricting the technology's implementation. In response, this paper proposes an imbalanced data classification method based on a hybrid parameter scaling weight optimization mechanism. First, a loss weight calculation model is constructed based on category proportion, category sparsity, and category diversity. Second, the loss weight calculation model is designed by integrating the focal loss function's ability to focus on hard samples with the cross-entropy loss function's global gradient stability advantage. Then, at the early stages of training, the model proposed in this paper aligns sensitivity to imbalanced categories and constructs a low-computational-demand hybrid parameter scaling weight optimization mechanism. Experimental results show that, compared with the best-performing baseline methods, the proposed method significantly improves both accuracy and macro-F1 score on the traffic accident disability grade dataset. It can effectively enhance the classification performance of minority grade categories in imbalanced data and help improve the accuracy of automated appraisal in judicial identification of traffic accident disability grades.

Branch-duct intraductal papillary mucinous neoplasms (BD-IPMNs) are pancreatic cystic lesions originating from the pancreatic ducts, characterized by mucin production and progressive ductal dilation. They exhibit a wide spectrum of biological behavior, ranging from indolent lesions to entities with significant malignant potential. Although the 2024 Kyoto guidelines define worrisome features (WF) and high-risk stigmata (HRS) to support risk stratification and clinical management, predicting disease progression remains challenging. In this retrospective study, we investigated whether MRI-based radiomic analysis could identify, at the time of initial imaging, patients with BD-IPMNs who subsequently develop WF or HRS according to 2024 Kyoto guidelines. A total of 194 adult patients who underwent at least two MRI examinations between January 2011 and March 2025 were included, with a median follow-up of 53months; progression was observed in 28.3% of patients, involving only some WF/HRS. Radiomic analysis included manual lesion segmentation, extraction of 107 features (shape, first- and second-order), and selection via LASSO within a weighted logistic regression framework to address class imbalance, using fivefold cross-validation; model performance was assessed with AUC and precision-recall metrics to account for skewed class distribution. After statistical analysis, nine shape-related features were found to be significant and a LASSO-based radiomic model, incorporating five features, was constructed. The model achieved an area under the curve (AUC) of 0.70 (95% CI 0.62-0.79). These results suggest that MRI-based radiomics may represent a valuable noninvasive tool for early risk stratification, predicting progression according to clinical-radiological criteria and potentially supporting personalized management of patients with BD-IPMNs. However, this study presents some limitations, including the retrospective design, the relatively small sample size and possible variability due to the use of multiple MRI scanners from different vendors. Prospective and multicentric studies with standardized imaging protocols are necessary to validate these findings and assess the added value of integrating radiomic data with clinical, histopathological and molecular information.

Genetic evaluation of common and novel resilience indicators under different milk recording frequencies in Holstein cattle.

The proliferation of distributed network environments and the Internet of Things (IoT) has increased the need for privacy-preserving intrusion detection systems capable of operating effectively under heterogeneous and non-independent and identically distributed (non-IID) data conditions. This paper proposes SecureTrust-FL, a trust-aware federated learning framework for privacy-preserving intrusion detection. The framework integrates Federated Learning, Blockchain-based Trust Management, Differential Privacy, FGSM-based Adversarial Learning, and Zero-Trust Security principles to support secure collaborative learning without requiring raw data sharing among participating entities. The framework is evaluated using three benchmark intrusion detection datasets, namely CICIDS2017, UNSW-NB15, and BoT-IoT, which are treated as independent federated clients. Experimental results demonstrate that the proposed framework achieves an overall Accuracy of 92.91% ± 0.45%, Balanced Accuracy of 93.25% ± 0.43%, Macro F1-Score of 92.89% ± 0.45%, and AUC-ROC of 95.50% ± 0.40% across heterogeneous datasets. The results indicate that the federated model can effectively learn from distributed and heterogeneous data while preserving data privacy. Further analysis reveals the impact of class imbalance on intrusion detection performance, particularly in datasets containing skewed attack distributions, highlighting the importance of Balanced Accuracy and F1-Score in addition to overall Accuracy. Differential privacy experiments demonstrate the privacy-utility trade-off, where stronger privacy protection leads to a reduction in model performance. Adversarial robustness evaluation using FGSM perturbations also shows a noticeable decline in detection performance, indicating the need for stronger defense mechanisms against adversarial attacks. In addition, the trust ledger enhances transparency and accountability by monitoring client participation and recording the trust scores used during trust-weighted aggregation and maintaining trust records throughout the collaborative learning process. The results demonstrate that SecureTrust-FL provides an effective framework for privacy-preserving collaborative intrusion detection while integrating trust management, privacy protection, and secure federated learning within a unified architecture.

Identifying prognostic risk factors in metabolic dysfunction-associated steatotic liver disease (MASLD) remains critical. This study investigates the association of the pan-immune inflammation value (PIV) and systemic immune inflammation index (SII) with mortality in the MASLD population. This cohort study used data from the National Health and Nutrition Examination Survey 1999 to 2018, associated with the National Death Index records, including 6252 patients with MASLD. Natural logarithms of PIV and SII (lnPIV, lnSII) were analyzed due to right-skewed distributions. Kaplan–Meier survival curves and multivariate Cox proportional hazards models were employed to assess the relationships between lnPIV, lnSII, all-cause and cardiovascular disease (CVD) mortality. Restricted cubic splines were applied to explore nonlinear trends, and segmented Cox regression was used to analyze threshold effects. Model performance was evaluated using the concordance index, time-dependent receiver operating characteristic curves, and calibration plots. Subgroup analyses were conducted to assess the robustness of the findings across different populations. Over a median follow-up of 9.17 years, 1154 all-cause deaths and 373 CVD-related deaths were recorded. Compared with individuals in the lowest quartile (Q1), those in the highest quartile (Q4) of lnPIV and lnSII had multivariable-adjusted hazard ratios of 1.36 (95% confidence interval [CI], 1.08–1.70) and 1.28 (95% CI, 1.02–1.61) for all-cause mortality, and 1.59 (95% CI, 1.00–2.52) and 1.76 (95% CI, 1.18–2.62) for CVD mortality, respectively. Nonlinear associations were observed between lnPIV, lnSII, and all-cause mortality, with thresholds identified at lnPIV = 6.015 and lnSII = 6.342. No significant associations were detected (P > .05) below these thresholds, whereas a significant positive association was found above the thresholds (P < .001). In contrast, a linear relationship was observed between lnPIV, lnSII, and CVD mortality. Time-dependent receiver operating characteristic curves and calibration plots demonstrated good model discrimination and calibration. No significant interactions were found across most subgroups (P for interaction > .05). Elevated PIV and SII are associated with an increased risk of all-cause and CVD mortality in MASLD. A significantly higher hazard of all-cause mortality was observed when lnPIV and lnSII were above specific thresholds. They are potential tools that require external validation in patients with MASLD.

The financial burden of medical expenses for patients with bronchial and pulmonary malignancies in coal-abundant regions poses a significant challenge. This study examines the variations in medical costs among patients with diverse characteristics and analyzes the distribution of expenses across different healthcare settings, thereby laying the foundation for the rational management of medical expenditures. Additionally, this research evaluates the adequacy of local medical insurance systems in supporting patients with bronchial and pulmonary malignancies, offering insights to improve the equity of medical insurance coverage. A total of 4,930 hospitalized patients with bronchial and pulmonary malignant tumors were identified from the Changzhi Medical Insurance Center records between January 2018 and June 2022. After transforming the skewed quantitative data on total medical costs, we conducted a cost composition analysis and a single-factor analysis. A BP neural network was subsequently employed to identify the influencing factors and evaluate the medical security level. Among the total costs, the five highest sub-cost categories were drug costs (31.76%), diagnosis and treatment costs (22.10%), examination costs (21.10%), consumable costs (19.20%), and comprehensive medical service costs (3.45%). Gender, surgical intervention, patient transfer, and hospital level exerted statistically significant effects on total medical costs (P < 0.05); among these, the three most influential factors were length of hospital stay (0.245), hospital grade (0.208), and surgical intervention (0.165). The actual compensation ratio in secondary-level medical institutions (mean 78.81%) was higher than that observed in tertiary-level medical institutions (mean 68.85%). Drug and consumable costs accounted for the highest proportion of the total costs. It is an effective measure to reduce the economic burden of patients by controlling the use of drugs and consumables, effectively monitoring the proportion of drug expenditure and diagnosis and treatment costs, appropriately increasing the salary of nursing staff, reasonably shortening the length of hospital stay, and improving the fairness of medical insurance reimbursement rate.

On the collapse of an endemic reef-building coral species.

Twitter's "EDTWT" community constitutes a prominent online space for eating disorder (ED) discourse, yet large-scale computational characterization remains limited. To characterize EDTWT behavioral patterns, emotional dynamics, thematic structure, and develop automated content classification methods through computational analysis of a three-year dataset. We analyzed 48,341 tweets from 18,587 users (January 2022-February 2025). Analyses included engagement patterns, temporal dynamics, clinical keyword prevalence, multi-method sentiment analysis (TextBlob, VADER, clinical affect lexicons), topic modeling (LDA, NMF), and ensemble-based multi-label classification. User engagement followed a highly skewed distribution in which a small number of highly active users generated the majority of content (62.1% of users posted only once, while the top 10% produced 48.9% of all content), suggesting that a concentrated subset of users may warrant particular clinical attention. Temporal patterns showed a 7.5-fold difference in posting volume between Friday night peaks and Tuesday morning troughs, with consistent nocturnal peaks between 21:00 and 23:00. Clinical keywords appeared in 26.6% of tweets (body image 14.4%, restrictive eating 6.7%, recovery 4.3%). Sentiment was slightly positive (M=0.060, SD=0.282) with moderate subjectivity (M=0.298). Topic modeling revealed ten themes including calorie tracking (13.2%), Spanish (9.7%) and Polish (8.2%) subcommunities, and recovery discourse (9.4%). An ensemble of automated classifiers trained to categorize eating disorder content achieved strong performance (macro F1=0.753, κ=0.703, where higher values indicate better classification accuracy), outperforming a biomedical-domain language model by 21.4% on the least frequently occurring content categories. EDTWT exhibits complex heterogeneity with coexisting pro-ED content, recovery discourse, and culturally-specific subcommunities. The concentration of activity among a small number of highly active users enables efficient identification of individuals who may be at elevated risk and in need of targeted support. Automated classification enables scalable content monitoring for digital mental health surveillance and evidence-based platform moderation.

ABSTRACT Rockeries in Qing-dynasty imperial gardens are irregular stone masonry structures combining aesthetic scene composition with engineering attributes. Existing geometric characterization methods, mostly oriented towards regular Western stone masonry walls and operating at a single scale, are inadequate for resolving the multi-tier composition rules on the primary viewing facade of rockeries. This study develops a multi-scale geometric quantification framework for the primary viewing facade of rockeries, extracting six categories of geometric factors at three scales: the overall layout, the single-stone hierarchy, and the individual stacked stone. The framework is validated on the Wanfang Anhe rockery in Yuanmingyuan. The results show that: (1) the overall layout exhibits a clear host-and-guest order, with the Main Rockery dominating in both area and count (area ratio 2.07:1.11:1.00); (2) at the single-stone hierarchy, the Rockery Side serves as the aggregation core (60.47% by area, 72.56% by count), while the Rockery Top forms an abrupt scale transition through a few large stones; (3) individual stones display slender, irregular shapes, a right-skewed size distribution (75% ≤0.12 m2) and near-random orientation (0.06°–179.90°). The proposed framework transforms qualitative rockery-stacking principles into quantifiable descriptions and may provide a methodological basis for stone-selection references in localized stacked-stone replacement and for the digital preservation of traditional construction principles.

Altmetric Attention Score (AAS) for the year 2001–2025 is assessed for academic recognition of Massachusetts Institute of Technology (MIT) publications. The aim of the study is to assess how open access (OA) affects the visibility of MIT articles in the research community, measure altmetric attention on various social media platforms, and assess the impact of citations. “We analyzed 313 MIT-affiliated publications that each had an AAS ≥ 1,000. Across this set, the sum of AAS values was 95,123, and the publications collectively attracted over 1,28,085 Mendeley readers during the study period.” The findings show a noticeable rise in both research output and online visibility since 2010, with the greatest altmetric attention happening throughout the COVID-19 pandemic era (2020–2022). The AAS among the chosen publications varies between 1,000 and 19,042. This variation suggests a right-skewed distribution. It has a mean of 2,018 and a median of 1,438. Altmetric attention is most strongly correlated with Google Scholar (GS) citations (Spearman’s p = 0.580, p < 0.001), followed by Scopus (p = 0.496) and Web of Science (WoS) (p = 0.485). Overall, the results suggest that altmetric indicators complement traditional citation metrics by capturing early online engagement to identify broader research visibility.

This study aims to establish an optimal ratio of Mineral Oil (MO) and Neem Oil After Esterification (NOAE)to develop a Mixed Insulation (MXI) that effectively replaces the MO for oil-filled transformers and other oil-filled electrical equipments. The effective use of two-stage esterification technique using Neem Oil (NO) as a base oil and concentrated sulfuric acid (H2SO4) and potassium hydroxide (KOH) as acid and base catalysts had been adopted to transform NO into NOAE. Further, the FAME formation is confirmed using the 1H-NMR. By lowering the impact of MO in oil-cooled transformers, the proposed MXI's dielectric properties are critically analysed using its AC breakdown voltage (BDV) in sphere-sphere and point-plane, kinematic viscosity (KVIS), corona inception voltage (CIV) and interfacial tension (IFT), and other dielectric properties. The MXI proportions are subject to accelerated thermal ageing in the presence of a pressboard and copper plate. The changes in their dielectric properties are reported along with the information about the material characteristics using an FTIR. Further, the validation of the samples is verified using Weibull distribution and Kurtosis and Skewness values. The current study on MXI is given credence by the discovery that sample X30, or the optimal MO: NOAE blending ratio, is 70:30. The X30's dielectric properties showed a notable level of durability after ageing, which helped to explain its high resistance and low conductivity.

Children with elevated low-density lipoprotein cholesterol (LDL-C) levels, suspected familial hypercholesterolemia (FH), and elevated lipoprotein(a) (Lp(a)) levels are considered to have a particularly high lifetime risk of atherosclerotic cardiovascular disease. Nevertheless, the distribution of Lp(a) levels among children with elevated LDL-C levels remains unclear. This study aimed to clarify the distribution of Lp(a) levels and their association with pathogenic FH variants in children with elevated LDL-C levels. A total of 97 children with LDL-C levels ≥ 140 mg/dL suspected of having FH who underwent genetic testing at Kagawa University Hospital between January 2018 and May 2025 were analyzed. Lp(a) levels were measured using the Lp(a) Latex "DAIICHI" assay and converted from mass units (mg/dL) to molar units (nmol/L) using the calibration-based conversion formula. Clinical and lipid parameters were compared according to Lp(a) levels (≥ 105 vs. ＜105 nmol/L) and the presence of pathogenic FH variants. Lp(a) levels exhibited a right-skewed distribution, with a median of 57.6 nmol/L (15.9 mg/dL) and an interquartile range of 24.0-93.4 nmol/L (7.0-25.4 mg/dL), and 21.6% of children had levels ≥ 105 nmol/L. Pathogenic FH variants were identified in 44 children. No significant differences were observed in either clinical or lipid parameters according to Lp(a) levels (≥ 105 vs. ＜105 nmol/L) or the presence of pathogenic FH variants (P = 0.672). Regardless of the presence of pathogenic FH variants, approximately 20% of the children had Lp(a) levels ≥ 105 nmol/L. These findings emphasize the importance of early lipid management and suggest that Lp(a) measurement may contribute to future cardiovascular risk stratification.

To address Cronbach's longstanding call to unify experimental and correlational psychology, hierarchical factor models (HFMs) have emerged as a promising approach to estimate individual differences while accounting for trial-level noise. As formalized by Rouder et al. (2025), this framework assumes Gaussian response times and operates in raw units. Here, we generalize this framework by introducing the standardized generalized HFM (GenHFM), which incorporates two natural extensions: Explicitly modeling asymmetric distributions and implementing a fully standardized latent structure. First, modeling asymmetry allows us to capture the true shape of response times and, crucially, quantify the bias introduced by fitting Gaussian HFMs to skewed data. Second, the standardized parameterization ensures transparent prior specification and facilitates the analysis of factor loadings, enabling researchers to directly compare the degree of association between experimental effects and common factors, regardless of their original metrics. Consequently, this allows researchers to evaluate the discriminative capacity of each task regarding the latent process, mirroring the assessment of construct validity in classical psychometrics within experimental settings. We conducted a simulation study showing that Gaussian HFMs can underestimate true correlations by up to 50%, whereas GenHFM yields unbiased and more efficient estimates. Finally, we reanalyze data from two published studies using executive control tasks to illustrate the empirical application of these models. Furthermore, we show that GenHFM achieves better predictive accuracy than alternative models typically used in the field. (PsycInfo Database Record (c) 2026 APA, all rights reserved).