Abstract
Wireless capsule endoscopy (WCE) has revolutionized the field of gastrointestinal examinations, being MedtronicTM WCE one of the most used in clinics. In those WCE videos, medical experts use RAPID READERTM tool to annotate findings in videos. However, the frame annotations are not available in an open format and, when exported, they have different resolutions and some annotated artefacts that make difficult their localization in the original videos. This difficult the use of WCE medical experts' annotations in the research of new computed-aid diagnostic (CAD) methods. In this paper, we propose a methodology to compare image similarities and evaluate it in a private MedtronicTM WCE SB3 video dataset to automatically identify the annotated frames in the videos. We used state-of-the-art pre-trained convolutional neural network (CNN) models, including MobileNet, InceptionResNetv2, ResNet50v2, VGG19, VGG16, ResNet101v2, ResNet152v2, and DenseNet121, as frame features extractors and compared them with the Euclidean distance. We evaluated the methodology performance on a private dataset consisting of 100 WCE videos, totalling 905 frames. The experimental results showed promising performance. The MobileNet model achieved an accuracy of 94% for identifying the first match, while the top 5, top 10, and top 20 matches were identified with accuracies of 94%, 94%, and 98%, respectively. The VGG16 and ResNet50v2 models also demonstrated strong performance, achieving accuracies ranging from 88% to 93% for various match positions. These results highlight the effectiveness of our proposed methodology in localizing target frames and even identifying similar frames very use useful for training data-driven models in CAD research. The code utilized in this experiment is available on the Github† © 2024 The Author(s). Published by Elsevier B.V.

Abstract
Colorectal cancer is a leading health concern worldwide, with late detection being a primary challenge due to its often-asymptomatic nature. Routine examinations like colonoscopies play a pivotal role in early detection. This study harnesses the potential of Deep Learning, specifically convolutional neural networks, in enhancing the accuracy of polyp detection from medical images. Three distinct models, YOLOv5, YOLOv7, and YOLOv8, were trained on the PICCOLO dataset, a comprehensive collection of polyp images. The comparative analysis revealed YOLOv5's submodel S as the most efficient, achieving an accuracy of 92.2%, a sensitivity of 69%, an F1 score of 74% and a mAP of 76.8%, emphasizing the effectiveness of these networks in polyp detection.

Abstract
Artificial intelligence (AI) is playing a growing role today in several areas, especially in health, where understanding AI models and their predictions is extremely important for health professionals. In this context, Explainable AI (XAI) plays a crucial role in seeking to provide understandable explanations for these models. This article analyzes two different XAI approaches applied to analyzing gastric endoscopy images. The first, more conventional approach uses Grad CAM, while the second, even less explored but with great potential, is based on similarity-based explanations. This example-based XAI technique aims to provide representative examples to support the decisions of AI models. In this study, we compare these two techniques applied to two different models: one based on the VGG16 architecture and the other based on ResNet50, designed to classify images from the KVASIR-capsule database. The results reveal that Grad-CAM provided intuitive explanations only for the VGG16 model, while the similarity-based explanations technique provided consistent explanations for both models. We conclude that exploring other XAI techniques can be a significant asset in improving the understanding of the various AI models.

Abstract
This research addresses the significant challenge of automating the annotation of medical images, with a focus on capsule endoscopy videos. The study introduces a novel approach that synergistically combines Deep Learning and Content-Based Image Retrieval (CBIR) techniques to streamline the annotation process. Two pre-trained Convolutional Neural Networks (CNNs), MobileNet and VGG16, were employed to extract and compare visual features from medical images. The methodology underwent rigorous validation using various performance metrics such as accuracy, AUC, precision, and recall. The MobileNet model demonstrated exceptional performance with a test accuracy of 98.4%, an AUC of 99.9%, a precision of 98.2%, and a recall of 98.6%. On the other hand, the VGG16 model achieved a test accuracy of 95.4%, an AUC of 99.2%, a precision of 97.3%, and a recall of 93.5%. These results indicate the high efficacy of the proposed method in the automated annotation of medical images, establishing it as a promising tool for medical applications. The study also highlights potential avenues for future research, including expanding the image retrieval scope to encompass entire endoscopy video databases.

Abstract
Glaucoma is a condition that affects the optic nerve, with loss of retinal nerve fibers, increased excavation of the optic nerve, and a progressive decrease in the visual field. It is the leading cause of irreversible blindness in the world. Manual classification of glaucoma is a complex and time-consuming process that requires assessing a variety of ocular features by experienced clinicians. Automated detection can assist the specialist in early diagnosis and effective treatment of glaucoma and prevent vision loss. This study developed a deep learning model based on vision transformers, called ViT-BRSET, to detect patients with increased excavation of the optic nerve automatically. ViT-BRSET is a neural network architecture that is particularly effective for computer vision tasks. The results of this study were promising, with an accuracy of 0.94, an F1-score of 0.91, and a recall of 0.94. The model was trained on a new dataset called BRSET, which consists of 16,112 fundus images of patients with increased excavation of the optic nerve. The results of this study suggest that ViT-BRSET has the potential to improve early diagnosis through early detection of optic nerve excavation, one of the main signs of glaucomatous disease. ViT-BRSET can be used to mass-screen patients, identifying those who need further examination by a doctor.

Abstract
Building rehabilitation is a reality, and all phases of rehabilitation work need to be efficiently sustainable and promote healthy places to live in. Current procedures for assessing construction conditions are time-consuming, laborious and expensive and pose threats to the health and safety of engineers, especially when inspecting locations that are not easy to access. In the initial step, a survey of the condition of the building is carried out, which subsequently implies the elaboration of a report on existing pathologies, intervention solutions, and associated costs. This survey involves an inspection of the site (through photographs and videos). Also, biological growth can threaten the humans inhabiting the houses. The World Health Organization states that the most important effects are increased prevalences of respiratory symptoms, allergies and asthma, as well as perturbation of the immunological system. This work aims to alert to this fact and contribute to detecting and locating biological growth (BG) defects automatically in images of the facade of buildings. To make this possible, we need a dataset of images of building components with and without biological growths. At this moment, that database doesn't exist. So, we need to construct that dataset to use deep learning models in the future. This paper also identifies the steps to do that work and presents some real cases of building facades with BG and solutions to repair those defects. The conclusions and the future works are identified.