Abstract
Lung cancer is a significant global health concern, and accurate classification of lung nodules plays a crucial role in its early detection and treatment. This paper evaluates and compares the performance of Vision Transformer (ViT) and Convolutional Neural Network (CNN) models for lung nodule classification using the Pylung tool proposed in this work. The study aims to address the lack of research on ViT in lung nodule classification and proposes ViT as an alternative to CNN. The Lung Image Database Consortium and Image Database Resource Initiative (LIDC-IDRI) dataset is utilized for training and evaluation. The Pylung tool is employed for dataset preprocessing and comparison of models. Three models, ViT, VGG16, and ResNet50, are analyzed, and their hyperparameters are optimized using Optuna. The results show that ViT achieves the highest accuracy (99.06%) in nodule classification compared to VGG16 (98.71%) and ResNet50 (98.46%). The study contributes by introducing ViT as a model for lung nodule classification, presenting the Pylung tool for model comparison, and suggesting further investigations to improve the accuracy. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

Abstract
This study focuses on investigating different CNN architectures and assessing their effectiveness in classifying Diabetic Retinopathy, a diabetes-associated disease that ranks among the primary causes of adult blindness. However, early detection can significantly prevent its debilitating consequences. While regular screening is advised for diabetic patients, limited access to specialized medical professionals can hinder its implementation. To address this challenge, deep learning techniques provide promising solutions, primarily through their application in the analysis of fundus retina images for diagnosis. Several CNN architectures, including MobileNetV2, VGG16, VGG19, InceptionV3, InceptionResNetV2, Xception, DenseNet121, ResNet50, ResNet50V2, and EfficientNet (ranging from EfficientNetB0 to EfficientNetB6), were implemented to assess and analyze their performance in classifying Diabetic Retinopathy. The dataset comprised 3662 Fundus retina images. Prior to training, the networks underwent pre-training using the ImageNet database, with a Gaussian filter applied to the images as a preprocessing step. As a result, the Efficient-Net stands out for achieving the best performance results with a good balance between model size and computational efficiency. By utilizing the EfficientNetB2 network, a model was trained with an accuracy of 85% and a screening capability of 98% for Diabetic Retinopathy. This model holds the potential to be implemented during the screening stages of Diabetic Retinopathy, aiding in the early identification of individuals at risk.

Abstract
Prostate cancer is one of the most common types of cancer in men. The ISUP grade and Gleason Score are terms related to the classification of this cancer based on the histological characteristics of the tissues examined in a biopsy. This paper explains an approach that utilizes and evaluates pre-trained models such as ResNet-50, VGG19, and InceptionV3, regarding their ability to automatically classify prostate cancer and its severity based on images and masks annotated with ISUP grades and Gleason Scores. At the end of the training, the performance of each trained model is presented, as well as the comparison between the original and predicted data. This comparison aims to understand if this approach can indeed be used for a more automated classification of prostate cancer.

Abstract
Growing obesity has been a worldwide issue for several years. This is the outcome of common nutritional disorders which results in obese individuals who are prone to many diseases. Managing diet while simultaneously dealing with the obligations of a working adult can be difficult. Today, people have a very fast-paced life and sometimes neglect food choices. In order to simplify the interpretation of the Nutri-score labeling this paper proposes a method capable of automatically reading food labels with this format. This method is intended to support users when choosing the products to buy based on the letter identification of the label. For this purpose, a dataset was created, and a prototype mobile application was developed using a deep learning network to recognize the Nutri-score information. Although the final solution is still in progress, the reading module, which includes the proposed method, achieved an encouraging and promising accuracy (above 90%). The upcoming developments of the model include information to the user about the nutritional value of the analyzed product combining it's Nutri-score label and composition.

Abstract
This paper presents a study on the automated detection of landmarks in medical x-ray images using deep learning techniques. In this work we developed two neural networks based on semantic segmentation to automatically detect landmarks in x-ray images, using a dataset of 200 encephalogram images: the UNet architecture and the FPN architecture. The UNet and FPN architectures are compared and it can be concluded that the FPN model, with IoU=0.91, is more robust and accurate in predicting landmarks. The study also had the goal of direct application in a medical context of diagnosing the models and their predictions. Our research team also developed a metric analysis, based on the encephalograms in the dataset, on the type of Mandibular Occlusion of the patients, thus allowing a fast and accurate response in the identification and classification of a diagnosis. The paper highlights the potential of deep learning for automating the detection of anatomical landmarks in medical imaging, which can save time, improve diagnostic accuracy, and facilitate treatment planning. We hope to develop a universal model in the future, capable of evaluating any type of metric using image segmentation.

Abstract
Canine hip dysplasia (CHD) screening relies on accurate positioning in the ventrodorsal hip extended (VDHE) view, as even mild pelvic rotation can affect CHD scoring and impact breeding decisions. This study aimed to assess the association between pelvic rotation and asymmetry in obturator foramina areas (AOFAs) and to develop a computer vision model for automated AOFA measurement. In the first part, 203 radiographs were analyzed to examine the relationship between pelvic rotation, assessed through asymmetry in iliac wing and obturator foramina widths (AOFWs), and AOFAs. A significant association was found between pelvic rotation and AOFA, with AOFW showing a stronger correlation (R-2 = 0.92, p < 0.01). AOFW rotation values were categorized into minimal (n = 71), moderate (n = 41), marked (n = 37), and extreme (n = 54) groups, corresponding to mean AOFA +/- standard deviation values of 33.28 +/- 27.25, 54.73 +/- 27.98, 85.85 +/- 41.31, and 160.68 +/- 64.20 mm(2), respectively. ANOVA and post hoc testing confirmed significant differences in AOFA across these groups (p < 0.01). In part two, the dataset was expanded to 312 images to develop the automated AOFA model, with 80% allocated for training, 10% for validation, and 10% for testing. On the 32 test images, the model achieved high segmentation accuracy (Dice score = 0.96; Intersection over Union = 0.93), closely aligning with examiner measurements. Paired t-tests indicated no significant differences between the examiner and model's outputs (p > 0.05), though the Bland-Altman analysis identified occasional discrepancies. The model demonstrated excellent reliability (ICC = 0.99) with a standard error of 17.18 mm(2). A threshold of 50.46 mm(2) enabled effective differentiation between acceptable and excessive pelvic rotation. With additional training data, further improvements in precision are expected, enhancing the model's clinical utility.