Abstract

Abstract
A key feature of model finding techniques allows users to enumerate and explore alternative solutions. However, it is challenging to guarantee that the generated instances are relevant to the user, representing effectively different scenarios. This challenge is exacerbated in quantitative modelling, where one must consider both the qualitative, structural part of a model, and the quantitative data on top of it. This results in a search space of possibly infinite candidate solutions, often infinitesimally similar to one another. Thus, research on instance enumeration in qualitative model finding is not directly applicable to the quantitative context, which requires more sophisticated methods to navigate the solution space effectively. The main goal of this paper is to explore a generic approach for navigating quantitative solution spaces and showcase different iteration operations, aiming to generate instances that differ considerably from those previously seen and promote a larger coverage of the search space. Such operations are implemented in QAlloy - a quantitative extension to Alloy - on top of Max-SMT solvers, and are evaluated against several examples ranging, in particular, over the integer and fuzzy domains.

Abstract
Diabetic retinopathy (DR) is a leading cause of vision loss in working-age adults. While screening reduces the risk of blindness, traditional imaging is often costly and inaccessible. Artificial intelligence (AI) algorithms present a scalable diagnostic solution, but concerns regarding fairness and generalization persist. This work evaluates the fairness and performance of image-trained models in DR prediction, as well as the impact of disentanglement as a bias mitigation technique, using the diverse mBRSET fundus dataset. Three models, ConvNeXt V2, DINOv2, and Swin V2, were trained on macula images to predict DR and sensitive attributes (SAs) (e.g., age and gender/sex). Fairness was assessed between subgroups of SAs, and disentanglement was applied to reduce bias. All models achieved high DR prediction performance in diagnosing (up to 94% AUROC) and could reasonably predict age and gender/sex (91% and 77% AUROC, respectively). Fairness assessment suggests disparities, such as a 10% AUROC gap between age groups in DINOv2. Disentangling SAs from DR prediction had varying results, depending on the model selected. Disentanglement improved DINOv2 performance (2% AUROC gain), but led to performance drops in ConvNeXt V2 and Swin V2 (7% and 3%, respectively). These findings highlight the complexity of disentangling fine-grained features in fundus imaging and emphasize the importance of fairness in medical imaging AI to ensure equitable and reliable healthcare solutions.

Abstract
Visible-Near-Infrared (Vis-NIR) spectroscopy, spanning approximately 400 to 2500 nm, is an innovative technology with growing relevance for diagnostics performed at the point of care (POC). This review explores the potential of Vis-NIR in veterinary medicine, highlighting its advantages over complex techniques like Raman and Fourier transform infrared spectroscopy (FTIR) by being rapid, non-invasive, reagent-free, and compatible with miniaturized, portable devices. The methodology involves directing a broadband light source, often using LEDs, toward the sample (e.g., blood, urine, faeces), collecting spectral information related to molecular vibrations, which is then analyzed using chemometric methods. Successful veterinary applications include hemogram analysis in dogs, cats, and Atlantic salmon, and quantifying blood in ovine faeces for parasite detection. Key limitations include spectral interference from strong absorbers like water and hemoglobin, and the limited penetration depth of light. However, combining Vis-NIR with Self-Learning Artificial Intelligence (SLAI) is shown to isolate and mitigate these multi-scale interferences. Vis-NIR spectroscopy serves as an important complement to centralized laboratory testing, holding significant potential to accelerate clinical decisions, minimize stress on animals during assessment, and improve diagnostic capabilities for both human and animal health, aligning with the One Health concept.

Abstract

Abstract
Blending green hydrogen into gas networks is subject to strict quality regulation. As Europe diversifies its Liquefied Natural Gas supply, existing literature often ignores the extreme chemical variability of these sources by assuming a static composition. This study proposes a nonlinear steady-state optimization model, strictly adhering to European standards, to maximize hydrogen injection across eight real-world LNG profiles. Results reveal severe sensitivity to the carrier gas: lean gas (Trinidad and Tobago, USA) restricts maximum hydrogen integration to 8.89% and 11.40% by volume due to insufficient heavy hydrocarbons. Conversely, carbon-rich gas (Nigeria) allows up to 20.10%, although inert gases degrade this capacity. Ultimately, this 20.10% maximum blend yields only a 6.77% emission reduction. The imported chemical profile dictates absolute limits, proving that static universal hydrogen quotas are thermodynamically unachievable year-round without continuous dynamic quality tracking.