Abstract
Segmentation of the lungs in Computed Tomography (CT) is very challenging due to changes in lung shape, size, and parenchyma pattern, as well as differences in imaging acquisition protocols. As a consequence, these models may not be robust and may decrease their performance when deployed in a clinical setting. The Continual Learning paradigm holds great promise since learning models continually acquire incoming knowledge, having the ability to adapt to changing environments. In this work, experience replay with random sampling of past data was implemented, using the original CT images and the corresponding ground-truths. Data from four different institutions were used to develop the experiments, and the models were evaluated on a cross-cohort dataset. Using raw data, the goal was to study how the datasets and their imaging patterns were related and what impact the training datasets have on one another. The catastrophic forgetting effect diminished for almost all datasets. For two of the in-domain test datasets there was forward and backward transfer, results that could be linked to a possible similarity between them. A mean DSC of 0.94 was obtained across all datasets. The results showed how the similarity or disparity between data from different institutions can influence the performance of learning models. © 2025 IEEE.

Abstract
Agricultural productivity is increasingly threatened by plant diseases, which can spread rapidly and lead to significant crop losses if not identified early. Detecting plant diseases accurately in diverse and uncontrolled environments remains challenging, as most current detection methods rely heavily on lab-captured images that may not generalise well to real-world settings. This paper aims to develop models capable of accurately identifying plant diseases across diverse conditions, overcoming the limitations of existing methods. A combined dataset was utilised, incorporating the PlantDoc dataset with web-sourced images of plants from online platforms. State-of-the-art convolutional neural network (CNN) architectures, including EfficientNet-B0, EfficientNet-B3, ResNet50, and DenseNet201, were employed and fine-tuned for plant leaf disease classification. A key contribution of this work is the application of enhanced data augmentation techniques, such as adding Gaussian noise, to improve model generalisation. The results demonstrated varied performance across the datasets. When trained and tested on the PlantDoc dataset, EfficientNet-B3 achieved an accuracy of 73.31%. In cross-dataset evaluation, where the model was trained on PlantDoc and tested on a web-sourced dataset, EfficientNet-B3 reached 76.77% accuracy. The best performance was achieved with the combination of the PlanDoc and web-sourced datasets resulting in an accuracy of 80.19% indicating very good generalisation in diverse conditions. Class-wise F1-scores consistently exceeded 90% for diseases such as apple rust leaf and grape leaf across all models, demonstrating the effectiveness of this approach for plant disease detection.

Abstract
Solar energy is a compelling and growing avenue for transitioning towards reliance on sustainable and environmentally friendly energy sources. However, the performance of individual photovoltaic (PV) panels is vulnerable to defects inflicted by weather exposure. The industry currently addresses these challenges by conducting manual inspections at each PV installation, a process that is highly time-consuming, financially burdensome, difficult to scale, prone to human error, and sometimes unfeasible due to topographical constraints. This study aims to develop a solution that detects these defects in real time resorting to an Autonomous Aerial Vehicle (AAV) equipped with for a thermal and a visual sensor. This paper contributes three major advancements to PV defect detection: a robust, standardized dataset built following IEC TS 62446-3 specifications with annotations from a real PV power plant, a novel multi-spectral approach that leverages early fusion of visual and thermal data captured via an AAV, and benchmark performance metrics for defect detection using state-of-the-art AI models. Our implementation uses Real-Time DEtection TRansformer (RT-DETR) and You Only Look Once (YOLO) models, achieving a mean Average Precision@50 (mAP) of 94% for RT-DETR and a mAP@50 of nearly 95% for YOLOv11. These results demonstrate the effectiveness of our approach in real-world settings, addressing a significant operational challenge in PV plant maintenance while establishing new performance benchmarks for automated defect detection in industrial solar installations.

Abstract
It is largely accepted that technology creates a digital divide in a development context. Generally, to face the digital divide, countermeasures such as capacity building, knowledge and know-how transfer, and beneficiaries' involvement are applied to accompany technology deployment for development projects. Although the efficiency of such countermeasures is relative and debatable, they doubtlessly contribute to create an ecosystem where hope is permitted, for developing countries, to catch up and to be successful in their digital transformation process. With the advent of artificial intelligence (AI) and its massive worldwide promotion, such hope does not seem to be allowed anymore in developing and less-developed countries. AI technologies are designed and developed for technologically advanced environments in wealthy countries, and it has been shown that they have the potential to exacerbate problems in less-wealthy nations. In this article, it is shown that AI technology is intrinsically digital divide pro, and that there is no possible countermeasure against its potentially devastating effects on international development. This leads to a substantial concern that progress in AI technologies and the pressure to adopt them may increase inequalities both between and within countries, in ways which counteract the overall purpose of development. We call this new unbeatable form of digital divide the 'digital divide (2.0)', and we argue that AI is a perfect example of technologies that create and consolidate it.

Abstract
Algorithms based on computer vision are commonly used in pre-operative procedures to help health professionals detecting blood vessels, which is also the case with the Deep Inferior Epigastric Perforators (DIEPs). These blood vessels are essential to produce a viable autologous DIEP flap, and the analysis of characteristics such as their location, diameter and course is essential to ensure the success of surgeries. This analysis is made by a team of radiology technicians and then validated by a surgeon, making it a complex process that can take up to 2 hours. The proposed algorithm called Assisted Vascular Analysis (AVA) was developed to ensure a faster alternative to the conventional methods, using automation to identify structures of interest such as the skin, umbilicus and fascia, while also requiring minimum input from the user to segment each DIEP (2 points for the subcutaneous portion and 2 for the intramuscular portion). The AVA feasibility tests where conducted using 6 Computed Tomography Angiographies (CTAs), with a total of 28 DIEP calibers obtained during surgery (ground truths) from patients that underwent breast reconstruction with a DIEP flap. The algorithm was evaluated for its capability to segment the DIEPs and measure their caliber, comparing the results with the ground truth calibers and the manual mapping done by the radiology technicians. The Root Mean Square Error (RMSE) metric shows that the calibers obtained by the AVA algorithm (0.57 millimeters) and the radiology technicians (0.46 millimeters) are very similar, with the radiology technicians gaining a smaller edge of 0.11 millimeters. These results are very promising, since the errors are inferior to the average image resolution (0.88 millimeters). It was also demonstrated that the AVA algorithm is a faster alternative to manual segmentation, taking around 10 minutes to fully analyze each CTA, while the radiology technicians takes around 1 hour to do the DIEP mapping and caliber measurements. In conclusion, AVA is a validated algorithm to segment DIEP vessels and a faster alternative compared with conventional methods. © 2025 IEEE.

Abstract
This study presents the development and assessment of a mobile application - the WalkingPAD app - aimed at promoting adherence to physical exercise among patients with Peripheral Arterial Disease (PAD). The assessment of adherence, acceptability, and usability was performed using mixed methods. Thirty-eight patients participated in the study with a mean age of 63.4 years (SD = 6.8). Thirty patients used the application for three months, responded to a semi-structured interview, and completed a task test and the System Usability Scale (SUS, ranging from 0 to 100). The application's adherence rate was 73%. When patients were asked about their reasons for using the app, the main themes that emerged were motivation, self-monitoring, and support in fulfilling a commitment. The average SUS score was 82.82 (SD = 18.4), indicating high usability. An upcoming version of the WalkingPAD app is expected to redesign both tasks - opening the app and looking up the walking history - which were rated as the most difficult tasks to accomplish. The new version of the WalkingPAD app will incorporate participants' comments and suggestions to enhance usability for this population.