Abstract
This study was initiated at a time of unprecedented uncertainty, as lecturers and educational institutions across the world tried to manage the move to online education as a result of the global COVID-19 pandemic. It started with lecturers' perspectives of their performance during that time to identify innovative teaching strategies beyond the priority of emergency teaching. The main goal was to identify the occurrence of more permanent changes in Higher Education after the pandemic. The research was based on a qualitative approach where faculty members were interviewed about their activities before, during and after lockdown periods. Data collected was analysed with the help of an algorithm based on Artificial Intelligence. Ultimately, it was possible to gather and evaluate practical solutions related to hybrid learning in Europe, Australia, and New Zealand, leading to recommendations for stakeholders in Higher Education.

Abstract
The integration of electric vehicles is paramount to the electrification of the transport sector, supporting the energy transition. The charging process of electric vehicles can be perceived as a controllable load and targeted with price or incentive-based programs. Demand-side management can optimize charging station performance and integrate renewable energy generation through electrical energy storage. Data flowing through charging stations can be used in computational approaches to solve open challenges and create new services, such as a dynamic pricing strategy, where the charging tariff depends on operating conditions. This work presents a data-driven service that optimizes day-ahead charging tariffs with a bilevel optimization problem and develops a case study around a large-scale pilot. The impact of photovoltaics and battery storage on the dynamic pricing scheme was assessed. A dynamic pricing strategy was found to benefit self-consumption and self-sufficiency of the charging station, increasing over 4 percentage points in some cases.

Abstract
With industry evolution and the development of Industry 4.0, manufacturers are trying to leverage it and find a way to increase productivity. Digital Twins (DT) technologies allow them to achieve this objective and revolutionize Product Life-cycle Management as they provide real-time information and insights for companies, allowing real-time product monitoring. Virtual Reality (VR) is a technology that permits users to interact with virtual objects in immersive environments; even under constant development, VR has proven efficient and effective in enhancing training. DT integration into immersive VR environments is constantly developing, with many challenges ahead. This study aims the development of an immersive virtual world for training integrated with DT technologies to handle all users' input using the simulator. Those were subject to a performance evaluation to understand how the application handles different input types, which confirmed the viability and reliability of this integration.

Abstract
Coronary artery calcium is a good indicator of coronary artery disease and can be used for cardiovascular risk stratification. Over the years, different deep learning approaches have been proposed to automatically segment coronary calcifications in computed tomography scans and measure their extent through calcium scores. However, most methodologies have focused on using 2D architectures which neglect most of the information present in those scans. In this work, we use a 3D convolutional neural network capable of leveraging the 3D nature of computed tomography scans and including more context in the segmentation process. In addition, the selected network is lightweight, which means that we can have 3D convolutions while having low memory requirements. Our results show that the predictions of the model, trained on the COCA dataset, are close to the ground truth for the majority of the patients in the test set obtaining a Dice score of 0.90 +/- 0.16 and a Cohen's linearly weighted kappa of 0.88 in Agatston score risk categorization. In conclusion, our approach shows promise in the tasks of segmenting coronary artery calcifications and predicting calcium scores with the objectives of optimizing clinical workflow and performing cardiovascular risk stratification.

Abstract
Training in a virtual environment can augment the current methods of professional's training, preparing them better for possible situations in the field of work while taking advantage of Virtual Reality (VR) benefits. This paper proposes a cost-effective immersive VR platform designed in real-context usage, consisting of an authoring tool that permits the creation and manipulation of training courses and the execution of these courses in an immersive environment. Accomplishing a good training experience in an immersive simulation requires an equilibrium between the simulator performance and the virtual world aesthetics quality. Thus, in addition to presenting the development of the proposed training platform based on Unity technologies, this paper describes an objective performance evaluation of a virtual training scene using the different render pipelines and across immersive and non-immersive setups. Results confirmed the platform's viability and revealed that the rendering pipeline should be defined according to the display device used.

Abstract
Case-based explanations are an intuitive method to gain insight into the decision-making process of deep learning models in clinical contexts. However, medical images cannot be shared as explanations due to privacy concerns. To address this problem, we propose a novel method for disentangling identity and medical characteristics of images and apply it to anonymize medical images. The disentanglement mechanism replaces some feature vectors in an image while ensuring that the remaining features are preserved, obtaining independent feature vectors that encode the images' identity and medical characteristics. We also propose a model to manufacture synthetic privacy-preserving identities to replace the original image's identity and achieve anonymization. The models are applied to medical and biometric datasets, demonstrating their capacity to generate realistic-looking anonymized images that preserve their original medical content. Additionally, the experiments show the network's inherent capacity to generate counterfactual images through the replacement of medical features.