Continuous glucose monitors (CGM) have improved the management of patients with type 1 diabetes (T1D), with glucose oxidase (GOx)-based sensors being the most used. However, they are potentially subject to both electrochemical and enzymatic interferences, including those related to changes of pH. The objective of this study is to investigate the effect of ethanol, given as beer along with a mixed meal, on the accuracy of a commercial GOx-CGM. Data from 12 T1D participants in a randomized crossover trial to evaluate the effect of meal composition and alcohol consumption on postprandial glucose concentration were used. Absolute error (AE) and mean absolute relative difference (MARD) were calculated. The differences between the alcohol and nonalcohol scenarios were assessed using the Mann–Whitney U and Wilcoxon signed-rank tests. The AE in the alcohol study was low, but significantly greater as compared to the study without alcohol (p-value = 0.0418). The MARD was numerically but not significantly greater. However, both variables were greater at pH < 7.36 and significantly affected by time only in the alcohol arm. In T1D, alcohol consumption affects the accuracy of a GOx-CGM. This effect could be at least partially related to the ethanol-induced changes in pH.

Recent studies have highlighted the potential of Large Language Models (LLMs) to become experts in specific areas of knowledge through the utilization of techniques that enhance their context. Nevertheless, an interesting and underexplored application in the literature is the creation of an LLM that specializes in research projects, as it could streamline the process of project ideation and accelerate the advancement of research initiatives. In this regard, the aim of this work is to develop a tool based on LLM technology capable of assisting the employees of technology centers in answering their queries related to research projects funded under the Horizon 2020 program. By facilitating the identification of suitable funding calls and the formation of consortia with partners meeting specific requirements, tasks that are traditionally time-intensive, the proposed tool has the potential to improve operational efficiency and enable technology centers to allocate their resources more effectively. To improve the model’s baseline performance, context extension techniques such as Retrieved Augmented Generation (RAG) and prompt engineering were explored. Specifically, different RAG approaches and configurations, along with a specialized prompt, were tested on the LLaMA 3 70B model, and their results were compared to those obtained without context extension. The proposed evaluation metrics, which aligned with human judgment while maintaining objectivity, revealed that RAG systems outperformed the standalone LLaMA 3 70B, achieving a rate of optimal responses of up to 46% compared to 0% for the baseline model. These findings emphasize that integrating RAG and prompt engineering pipelines into LLMs can address key limitations, such as generating accurate and informative answers. Moreover, this study demonstrates the practical feasibility of leveraging advanced LLM configurations to support research-driven organizations, highlighting a pathway for the further development of intelligent tools that enhance productivity and foster innovation in the research domain.

Depopulation poses a territorial challenge in Colombia, exacerbated by the long-term effects of armed conflict, inequality, and rural poverty, which have resulted in uneven demographic and development patterns. This study examines the factors influencing territories at risk of demographic decline in Colombia. This is a territorial phenomenon whose implications vary across regions, and are shaped by Colombia’s social, cultural, and geographic diversity. A model that estimates depopulation risk by incorporating economic variables, conflict, poverty, and rurality was developed using machine learning techniques. The Support Vector Machine approach achieved an f1-score of 78.31%, demonstrating its effectiveness as a predictive tool. The framework emphasizes a regional perspective, enabling the identification and comparison of differences between regions through municipal-level variables that capture territorial characteristics. The model provides a practical tool for predicting and preventing depopulation risks, allowing policymakers to design tailored strategies for regions identified as vulnerable. This perspective is particularly relevant in Colombia, where departments are the key administrative units for regional planning and policy design. However, the proposed approach is not limited to the Colombian context. It can be adapted to other national contexts with varying territorial divisions, offering a flexible framework for analysing territorial heterogeneity and regional disparities. By linking data-driven modeling with regional development analysis, the model contributes to the design of informed strategies for addressing demographic decline and its long-term implications for regional development. Therefore, this study contributes to broader discussions on demographic change, territorial inequality, and the design of informed public policies in diverse socio-spatial contexts.

The beneficial relationship between Green Infrastructure and human wellbeing has been largely supported in literature. This work evaluates this beneficial relationship in two population groups (adults and minors). Health data for both groups was retrieved from reports and questionnaires, whereas data related to Green Infrastructure was obtained from the Corine Land Cover project. The naturalness of the cities was associated with fewer negative behavioural changes in minors: the rate of negative changes was inversely correlated with the Land Cover Naturalness index (r = -0.84, p = 0.039) and, although not significantly, with Green Natural Area (r = -0.74, p = 0.093), whereas it was positively correlated with Green Urban Area (r = 0.98, p < 0.001). In adults, the greater Green Area Density, the more optimistic inhabitants are about their health status perception (p-value = 0.031). In conclusion, natural spaces, as opposed to artificial green urban areas, may be associated with fewer negative behaviors in minors. Besides, a higher density of green areas per inhabitant was associated with better quality of life and self-perception of health and with lower sedentarism and obesity in adults.

With the goal of facilitating efficient access to data to support the development of new R&D&I initiatives, this paper presents an information retrieval RAG-based system for projects funded by the Horizon 2020 program domain through multiple configurations and specific adaptations, considering CPU-only hardware demand. This work is an extension of a previous proposal in order to address the limitations when moving from exploratory prototypes to robust, large-scale experimentation. Additionally, we propose an automatic evaluation pipeline that enables large-scale, reproducible comparison of different configurations by analyzing generated answers and automating the scoring process. A methodology structured into three main phases guided the development of the system. First, project embeddings were generated using vector representation techniques, segmenting documents to compare models in terms of semantic quality. Second, relevant fragment retrieval was performed using language models to assess their relevance to a query. Third, system performance was evaluated using language model-based evaluators, automating the process. Additionally, an in-depth analysis identified evaluators that best approximate human judgment, comparing their results with metrics from the reference study. This work contributes to the State of the Art by proposing improvements in information retrieval systems.