Abstract
Background/Objectives: Obesity is a major contributor to cardiovascular disease, yet traditional risk assessment methods may overlook behavioral and circadian influences that modulate metabolic health. Chronotype, physical activity, sleep quality, eating speed, and breakfast habits have been increasingly associated with cardiometabolic outcomes. This study aims to evaluate the associations between these behavioral factors and both anthropometric and biochemical markers of cardiovascular risk among obese candidates for bariatric surgery. Methods: A cross-sectional study was conducted in a sample of 286 obese adults (78.3% females, mean 44.3 years, SD = 10.8, mean BMI = 42.5 kg/m2, SD = 6.2) followed at a central Portuguese hospital. Chronotype (reduced Morningness-Eveningness Questionnaire), sleep quality (Pittsburgh Sleep Quality Index), physical activity (Godin-Shephard Questionnaire), eating speed, and breakfast skipping were assessed. Cardiovascular risk markers included waist-to-hip ratio (WHR), waist-to-height ratio, A Body Shape Index (ABSI), Body Roundness Index, atherogenic index of plasma (AIP), triglyceride-glucose index (TyG), and homeostatic model assessment for insulin resistance (HOMA-IR). Results: Men exhibited significantly higher WHR, ABSI, HOMA-IR, TyG, and AIP. Eveningness was associated with higher insulin (r = -0.168, p = 0.006) and HOMA-IR (r = -0.156, p = 0.011). Poor sleep quality was associated with higher body fat mass (r = 0.151, p = 0.013), total cholesterol (r = 0.169, p = 0.005) and LDL cholesterol (r = 0.132, p = 0.030). Faster eating speed was associated with a higher waist circumference (r = 0.123, p = 0.038) and skeletal muscle mass (r = 0.160, p = 0.009). Conclusions: Male sex, evening chronotype, and poor sleep quality were associated with more adverse cardiometabolic profiles in individuals with severe obesity. These findings support the integration of behavioral and circadian factors into cardiovascular risk assessment strategies.

Abstract
Electronic Health Records (EHRs) contain vast amounts of unstructured narrative text, posing challenges for organization, curation, and automated information extraction in clinical and research settings. Developing effective annotation schemes is crucial for training extraction models, yet it remains complex for both human experts and Large Language Models (LLMs). This study compares human- and LLM-generated annotation schemes and guidelines through an experimental framework. In the first phase, both a human expert and an LLM created annotation schemes based on predefined criteria. In the second phase, experienced annotators applied these schemes following the guidelines. In both cases, the results were qualitatively evaluated using Likert scales. The findings indicate that the human-generated scheme is more comprehensive, coherent, and clear compared to those produced by the LLM. These results align with previous research suggesting that while LLMs show promising performance with respect to text annotation, the same does not apply to the development of annotation schemes, and human validation remains essential to ensure accuracy and reliability. © 2025 Copyright for this paper by its authors.

Abstract

Abstract

Abstract
This report describes the ICDAR 2025 Competition on Automatic Classification of Literary Epochs (ICDAR 2025 CoLiE), which consisted of two tasks focused on automatic prediction of the time in which a book was written (date of first publication). Both tasks comprised two sub-tasks, where a related fine-grained classification was addressed. Task 1 consisted of the identification of literary epochs, such as Romanticism or Modernism (sub-task 1.1), and a more precise classification of the period within the epoch (sub-task 1.2). Task 2 addressed the chronological identification of century (sub-task 2.1) or decade (sub-task 2.2). The compiled dataset and the reported findings are valuable to the scientific community and contribute to advancing research in the automatic dating of texts and its applications in digital humanities and temporal text analysis. © 2025 Elsevier B.V., All rights reserved.

Abstract