Abstract
This article describes a Kernel Principal Component Regressor (KPCR) to identify Auto Regressive eXogenous (ARX) Linear Parmeter Varying (LPV) models. The new method differs from the Least Squares Support Vector Machines (LS-SVM) algorithm in the regularisation of the Least Squares (LS) problem, since the KPCR only keeps the principal components of the Gram matrix while LS-SVM performs the inversion of the same matrix after adding a regularisation factor. Also, in this new approach, the LS problem is formulated in the primal space but it ends up being solved in the dual space overcoming the fact that the regressors are unknown. The method is assessed and compared to the LS-SVM approach through 2 Monte Carlo (MC) experiments. Every experiment consists of 100 runs of a simulated example, and a different noise level is used in each experiment,with Signal to Noise Ratios of 20db and 10db, respectively. The obtained results are twofold, first the performance of the new method is comparable to the LS-SVM, for both noise levels, although the required calculations are much faster for the KPCR. Second, this new method reduces the dimension of the primal space and may convey a way of knowing the number of basis functions required in the Kernel. Furthermore, having a structure very similar to LS-SVM makes it possible to use this method in other types of models, e.g. the LPV state-space model identification.

Abstract
The segmentation of retinal vessels in fundus images has been heavily focused in the past years, given their relevance in the diagnosis of several health conditions. Even though the recent advent of deep learning allowed to foster the performance of computer-based algorithms in this task, further improvement concerning the detection of vessels while suppressing background noise has clinical significance. Moreover, the best performing state-of-the-art methodologies conduct patch-based predictions. This, put together with the preprocessing techniques used in those methodologies, may hinder their use in screening scenarios. Thus, in this paper, we explore a fully convolutional setting that takes raw fundus images and allows to combine patch-based training with global image prediction. Our experiments on the DRIVE, STARE and CHASEDB1 databases show that the proposed methodology achieves state-of-the-art performance in the first and the last, allowing at the same time much faster segmentation of new images.

Abstract
The enhancement of tubular structures such as vessels in medical images has been addressed in the past, aiming for easier extraction and or visualization of such structures by professionals. Some literature methodologies propose vesselness measures whose design is motivated by local properties of vascular networks and how these influence the eigenvalues of the Hessian matrix. However, past work fails to combine properly the scale-space and neighborhood information, thus leading to the proposal of suboptimal vesselness measures. In this paper, we show that a shallow convolutional neural network is able to learn more optimal embedding spaces from the eigenvalue analysis at different scales, thus leading to a stronger vessel enhancement. Additionally, we also show that such a system maintains one of the biggest advantages of Hessian-based vesselness measures, which is the robustness to data with varying statistics. © 2019, Springer Nature Switzerland AG.

Abstract
Lung cancer has an increasing preponderance in worldwide mortality, demanding for the development of efficient screening methods. With this in mind, a binary classification method using Lung-RADS (TM) guidelines to warn changes in the screening management is proposed. First, having into account the lack of public datasets for this task, the lung nodules in the LIDC-IDRI dataset were re-annotated to include a Lung-RADS (TM)-based referral label. Then, a wide residual network is used for automatically assessing lung nodules in 3D chest computed tomography exams. Unlike the standard malignancy prediction approaches, the proposed method avoids the need to segment and characterize lung nodules, and instead directly defines if a patient should be submitted for further lung cancer tests. The system achieves a nodule-wise accuracy of 0.87 +/- 0.02.

Abstract
Cardiovascular disease is the leading cause of death around the world and its early detection is a key to improving long-term health outcomes. To detect possible heart anomalies at an early stage, an automatic method enabling cardiac health low-cost screening for the general population would be highly valuable. By analyzing the phonocardiogram (PCG) signals, it is possible to perform cardiac diagnosis and find possible anomalies at an early-term. Accordingly, the development of intelligent and automated analysis tools of the PCG is very relevant. In this work, the PCG signals are studied with the main objective of determining whether a PCG signal corresponds to a “normal” or “abnormal” physiological state. The main contribution of this work is the evidence provided that time domain features can be combined with features extracted from a wavelet transformation of PCG signals to improve automatic cardiac disease classification. We empirically demonstrate that, from a pool of alternatives, the best classification results are achieved when both time and wavelet features are used by a Support Vector Machine with a linear kernel. Our approach has obtained better results than the ones reported by the challenge participants which use large amounts of data and high computational power. © Springer Nature Switzerland AG 2019.

Abstract
Central serous chorioretinopathy is a retinal disease in which there is a leak of fluid into the subretinal space resulting in mild to moderate loss of visual acuity. Sequences of images from a fluorescein angiography exam are most of the times used for analyzing these leaks. This work presents a diagnostic aid method to detect and characterize the progression of fluid area along the exam, in order to provide a second opinion and increase the focus and the speed of analysis of the ophthalmologists. The method is based on a comparative approach by image subtraction between the late and early frames. The obtained segmentation results are quite promising with an average Dice coefficient of 0.801 +/- 0.106 for the training set and 0.774 +/- 0.106 for the test set.