Abstract
With the ever-increasing occurrence of skin cancer, timely and accurate skin cancer detection has become clinically more imperative. A clinical mobile-based deep learning approach is a possible solution for this challenge. Nevertheless, there is a major impediment in the development of such a model: the scarce availability of labelled data acquired with mobile devices, namely macroscopic images. In this work, we present two experiments to assemble a robust deep learning model for macroscopic skin lesion segmentation and to capitalize on the sizable dermoscopic databases. In the first experiment two groups of deep learning models, U-Net based and DeepLab based, were created and tested exclusively in the available macroscopic images. In the second experiment, the possibility of transferring knowledge between the domains was tested. To accomplish this, the selected model was retrained in the dermoscopic images and, subsequently, fine-tuned with the macroscopic images. The best model implemented in the first experiment was a DeepLab based model with a MobileNetV2 as feature extractor with a width multiplier of 0.35 and optimized with the soft Dice loss. This model comprehended 0.4 million parameters and obtained a thresholded Jaccard coefficient of 72.97% and 78.51% in the Dermofit and SMARTSKINS databases, respectively. In the second experiment, with the usage of transfer learning, the performance of this model was significantly improved in the first database to 75.46% and slightly decreased to 78.04% in the second. © 2020, The Author(s).

Abstract
EGFR and KRAS are the most frequently mutated genes in lung cancer, being active research topics in targeted therapy. The biopsy is the traditional method to genetically characterise a tumour. However, it is a risky procedure, painful for the patient, and, occasionally, the tumour might be inaccessible. This work aims to study and debate the nature of the relationships between imaging phenotypes and lung cancer-related mutation status. Until now, the literature has failed to point to new research directions, mainly consisting of results-oriented works in a field where there is still not enough available data to train clinically viable models. We intend to open a discussion about critical points and to present new possibilities for future radiogenomics studies. We conducted high-dimensional data visualisation and developed classifiers, which allowed us to analyse the results for EGFR and KRAS biological markers according to different combinations of input features. We show that EGFR mutation status might be correlated to CT scans imaging phenotypes; however, the same does not seem to hold for KRAS mutation status. Also, the experiments suggest that the best way to approach this problem is by combining nodule-related features with features from other lung structures.

Abstract
Overuse of antibiotics is causing the environment to become polluted with them. This is a major threat to global health, with bacteria developing resistance to antibiotics because of it. To monitor this threat, multiple antibiotic detection methods have been developed; however, they are normally complex and costly. In this work, an affordable, easy to use alternative based on digital colourimetry is proposed. Photographs of samples next to a colour reference target were acquired to build a dataset. The algorithm proposed detects the reference target, based on binarisation algorithms, in order to standardise the collected images using a colour correction matrix converting from RGB to XYZ, providing a necessary colour constancy between photographs from different devices. Afterwards, the sample is extracted through edge detection and Hough transform algorithms. Finally, the sulfonamide concentration is estimated resorting to an experimentally designed calibration curve, which correlates the concentration and colour information. Best performance was obtained using Hue colour, achieving a relative standard deviation value of less than 3.5%. © 2019, Springer Nature Switzerland AG.

Abstract
Breast Ultrasound has long been used to support diagnostic and exploratory procedures concerning breast cancer, with an interesting success rate, specially when complemented with other radiology information. This usability can further enhance visualization tasks during pre-treatment clinical analysis by coupling the B-Mode images to 3D space, as found in Magnetic Resonance Imaging (MRI) per instance. In fact, Lesions in B-mode are visible and present high detail when comparing with other 3D sequences. This coupling, however, would be largely benefited from the ability to match the various structures present in the B-Mode, apart from the broadly studied lesion. In this work we focus on structures such as skin, subcutaneous fat, mammary gland and thoracic region. We provide a preliminary insight to several structure segmentation approaches in the hopes of obtaining a functional and dependable pipeline for delineating these potential reference regions that will assist in multi-modal radiological data alignment. For this, we experiment with pre-processing stages that include Anisotropic Diffusion guided by Log-Gabor filters (ADLG) and main segmentation steps using K-Means, Meanshift and Watershed. Among the pipeline configurations tested, the best results were found using the ADLG filter that ran for 50 iterations and H-Maxima suppression of 20% and the K-Means method with $$K=6$$. The results present several cases that closely approach the ground truth despite overall having larger average errors. This encourages the experimentation of other approaches that could withstand the innate data variability that makes this task very challenging. © Springer Nature Switzerland AG 2020.

Abstract
Breast cancer imaging research has seen continuous progress throughout the years. Innovative visualization tools and easier planning techniques are being developed. Image segmentation methodologies generally have best results when applied to specific types of exams or sequences, as their features enhance and expedite those approaches. Particular methods have more purchase with the segmentation of particular structures. This is the case with diverse breast structures and the respective lesions on MRI sequences, over T1w and Dyn. The present study presents a methodology to tackle an unapproached task. We aim to facilitate the volumetric alignment of data retrieved from T1w and Dyn sequences, leveraging breast surface segmentation and registration. The proposed method revolves around Canny edge detection and mending potential holes on the surface, in order to accurately reproduce the breast shape. The contour is refined with a Level-set approach and the surfaces are aligned together using a restriction of the Iterative Closest Point (ICP) method. This could easily be applied to other paired same-time, volumetric sequences. The process seems to have promising results as average two-dimensional contour distances are at sub-voxel resolution and visual results seem well within range for the valid transference of other segmented or annotated structures. © Springer Nature Switzerland AG 2020.

Abstract