Abstract
The link between epicardial adipose tissue (EAT) and cardiovascular risk is well established, with EAT volume being strongly associated with inflammation, coronary artery disease (CAD) risk, and mortality. However, its EAT quantification is hindered by the time-consuming nature of manual EAT segmentation in cardiac computed tomography (CT). 300 non-contrast cardiac CT scans were collected and the pericardium was manually delineated. In a subset of this data (N = 30), manual delineation was repeated by the same operator and by a second operator. Two automatic methods were then used for pericardial segmentation: a commercially available tool, Siemens Cardiac Risk Assessment (CRA) software; and a deep learning solution based on a U-Net architecture trained exclusively with external public datasets (CardiacFat and OSIC). EAT segmentations were obtained through thresholding to [- 150,- 50] Hounsfield units. Pericardial and EAT segmentation performance was evaluated considering the segmentations by the first operator as reference. Statistical significance of differences for all metrics and segmentation methods was tested through Student t-tests. Pericardial segmentation intra-/interobserver variability was excellent, with the U-Net outperforming Siemens CRA (p < 0.0001). The intra- and interobserver agreement for EAT segmentation was lower with Dice Scores (DSC) of 0.862 and 0.775 respectively, while the U-Net and Siemens CRA obtained DSCs of 0.723 and 0.679 respectively. EAT volume quantification showed that the agreement between a human observer and the U-Net was better than that of two human observers (p = 0.0141), with a Pearson Correlation Coefficient (PCC) of 0.896 and a bias of - 2.83 cm(3) (below the interobserver bias of 9.05 cm3). The lower performances of EAT segmentation highlight the difficulty in segmenting this structure. For both pericardial and EAT segmentation, the deep learning method outperformed the commercial solution. While the segmentation performance of the U-Net solution was below interobserver variability, EAT volume quantification performance was competitive with human readers, motivating future use of these tools. Clinical trial number: NCT03280433, registered retrospectively on 2017-09-08.

Abstract
Interprofessional collaboration is an essential competency for healthcare professionals, and escape rooms have emerged as an innovative strategy to enhance teamwork and communication. The purpose of this scoping review was to identify and summarize how escape rooms are used in the teaching and enhancement of interprofessional collaboration skills. We conducted a scoping review following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses for Scoping Review (PRISMA-ScR) guidelines. A search of five databases, Scopus (R), Web of Science (R), CINAHL Complete (R), MEDLINE (R) and PsychINFO (R) was conducted for all articles until 1 January 2024. The review included 15 studies, mostly from the USA, involving a total of 2,434 participants across various healthcare professions. Key findings indicated significant improvements in group cohesion, communication, understanding of team roles, and interprofessional skills. Escape rooms can be an effective pedagogical tool in enhancing interprofessional competencies among healthcare students and professionals. Further research is needed to explore the sustainability of skills gained over time through escape rooms and to refine assessment methods.

Abstract
Stroke affects over 100 million people worldwide, and over two-thirds of survivors experience lasting upper-limb impairments, which significantly impact their quality of life. The global shortage of rehabilitation providers, who cannot attend to all patients who need it, creates an urgent, not yet answered, need for reliable and accessible rehabilitation innovations. Robotic rehabilitation has been emerging as an effective alternative to traditional physical therapy. This paper presents the development and evaluation of 2 degree-of-freedom exoskeleton, coupled to a collaborative robotic manipulator, which performs upper-limb rehabilitation. The system targets elbow flexion/extension and forearm pronation/supination, using two direct current brushless actuators. To accommodate a wide range of users, the mechanical design is modular and adjustable, allowing the rehabilitation of a broad range of arm lengths, while mechanical barriers prevent unsafe joint motions. Furthermore, limit switches ensure the movements are performed within safe values and an emergency button is also available for emergency stop. Safety assessment confirmed the actuators' performance and the integrity of the physical barriers. Three different rehabilitation modes were implemented: passive assist, active assist and active resist. Passive assistance tests achieved consistent trajectory tracking with a root mean square error of 4.85(o)& strns; for pronation/supination and 0.87 & strns;(o) for elbow flexion/extension, while maintaining smooth motion profiles with spectral arc length values of-1.603 and-1.56, respectively. Active resistance generated stable bidirectional torque across the full range of motion, reaching up to 1 Nm for forearm pronation/supination and 7 Nm for elbow flexion/extension. The adaptive active assistance strategy modified the assistance torque in real time according to the detected user performance. These findings establish a foundation for future clinical evaluation and real-world applications, with the system's modular design and multiple therapy modes showing potential to support diverse rehabilitation needs.

Abstract
Monitoring solar phenomena, such as sunspots and active regions, is crucial for ensuring astronaut safety, telecommunications reliability, and predicting terrestrial events like auroras. Traditional methods for detecting these phenomena have limitations in accuracy and baseline maintenance. This paper presents a novel deep learning object detection method that leverages multispectral image data from satellites to enhance the detection of sunspots and active regions. Utilizing images from the SDO satellite and annotations from the DeepSDO dataset, we constructed a new dataset composed of aligned observations from HMI Ic, AIA 211 & Aring;, and AIA 335 & Aring;. We adapted and developed a stock YOLOv5-based model capable of handling and fusing any number of input images. Two fusion methodologies, early and late fusion, and three different fusion modules-CatFuse (simple concatenation), CBAMC (CBAM-based module), and TransEnc (transformer encoder)-were implemented and tested. Statistically analysing the results via the Friedman test (p=0.05) revealed significant performance differences among the evaluated models, which were confirmed through pairwise Wilcoxon post-hoc tests. From the approaches tested, CatFuse with early fusion achieved significantly higher detection performance than the other models, with a mAP@0.5:0.95 of 0.52 and a mAP@0.5 of 0.94, an improvement of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta $$\end{document}=0.02-0.36 for mAP@0.5:0.95 and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta $$\end{document}=0.02-0.28 for mAP@0.5, depending on the baseline model. This result was marginally better than the best baseline (YOLOv5 with a single HMI image) and comparable to other state-of-the-art models, demonstrating a modest but consistent improvement of multispectral image fusion for this task.

Abstract
The typical offline protocol to evaluate recommendation algorithms is to collect a dataset of user-item interactions and then use a part of this dataset to train a model, and the remaining data to measure how closely the model recommendations match the observed user interactions. This protocol is straightforward, useful and practical, but it only provides snapshot performance. We know, however, that online systems evolve over time. In general, it is a good idea that models are frequently retrained with recent data. But if this is the case, to what extent can we trust previous evaluations? How will a model perform when a different pattern (re)emerges? In this paper we propose a methodology to study how recommendation models behave when they are retrained. The idea is to profile algorithms according to their ability to, on the one hand, retain past patterns - stability - and, on the other hand, (quickly) adapt to changes - plasticity. We devise an offline evaluation protocol that provides detail on the long-term behavior of models, and that is agnostic to datasets, algorithms and metrics. To illustrate the potential of this framework, we present preliminary results of three different types of algorithms on the GoodReads dataset that suggest different stability and plasticity profiles depending on the algorithmic technique, and a possible trade-off between stability and plasticity. We further discuss the potential and limitations of the proposal and advance some possible improvements. © 2026 Copyright held by the owner/author(s).

Abstract
As organizations navigate through complex and collaborative digital environments, Generative AI (GenAI) emerges as a transformative force for Knowledge Management (KM) processes. This paper highlights how GenAI technologies impact collaborative KM processes across individual, intraorganizational, and inter-organizational levels within the evolving paradigm of Industry 5.0 (i5.0). Through a literature review, the study explores how GenAI augments human cognition, enhances knowledge creation and sharing, and fosters organizational adaptability and innovation. The findings highlight GenAI's potential as cognitive partner, streamlining information flows, and improving decision-making across collaborative networks. However, challenges such as over-reliance, ethical risks, and the decline of critical human skills are also discussed. Furthermore, the paper identifies the evolution and gaps in current literature on Collaborative Networks (CNs) regarding the integration of AI technologies. It contributes to the ongoing discussion towards a socio-technical transformation while also providing an overview for rethinking collaboration and social strategies in the GenAI era.