Abstract
Chest radiography is one of the most ubiquitous medical imaging exams used for the diagnosis and follow-up of a wide array of pathologies. However, chest radiography analysis is time consuming and often challenging, even for experts. This has led to the development of numerous automatic solutions for multipathology detection in chest radiography, particularly after the advent of deep learning. However, the black-box nature of deep learning solutions together with the inherent class imbalance of medical imaging problems often leads to weak generalization capabilities, with models learning features based on spurious correlations such as the aspect and position of laterality, patient position, equipment and hospital markers. In this study, an automatic method based on a YOLOv3 framework was thus developed for the detection of markers and written labels in chest radiography images. It is shown that this model successfully detects a large proportion of markers in chest radiography, even in datasets different from the training source, with a low rate of false positives per image. As such, this method could be used for performing automatic obscuration of markers in large datasets, so that more generic and meaningful features can be learned, thus improving classification performance and robustness.

Abstract

Abstract

Abstract
Optical fiber gratings have long shown their sensing capabilities. One of the main challenges, however, is the interrogation method applied, since typical systems tend to use broadband light sources with optical spectrum analyzers, laser scanning units or CCD (Charged Coupled Device) spectrometers. The following paper presents the development of an interrogation system, which explores the temperature response of a multimode laser diode, in order to interrogate long period fiber gratings. By performing a spectral sweep along one of its rejection bands, a discrete attenuation spectrum is created. Through a curve fitting technique, the original spectrum is restored. The built unit, while presenting a substantially reduced cost compared with typical interrogation systems, is capable of interrogating along a 10 nm window with measurement errors reaching minimum values as low as 0.4 nm, regarding the grating central wavelength, and 0.4 dB for its attenuation. Given its low cost and reduced dimensions, the developed system shows potential for slow-changing field applications.

Abstract
This paper deals with the bi-level two-stage operation scheduling of hydrogen-based microgrids within a distribution system where the wind and solar generation and load demands are considered as uncertain variables. The distribution system is considered as a leader in the upper level and microgrids as followers in the lower level. Unlike previous approaches, the upper-level is within the day-ahead market and considered a deterministic problem, and the lower-level is considered a stochastic problem and consists of two stages. The first stage determines the purchasing power from the distribution system, while the second stage adjusts the outputs and power dispatch for any realizations of scenarios. This model is transformed from a bi-level to a linear single-level model by applying the Karush- Kuhn-Tucker (KKT) optimally conditions, strong duality, and Fortuny-Amat methods. Several comparisons have been carried out regarding the single clearing price for all microgrids or separate prices for each microgrid. Furthermore, power exchange and dispatch in the distribution system are investigated under the mentioned frameworks.

Abstract
Convolutional neural networks (CNN) and transfer learning are receiving a lot of attention because of the positive results achieved on image recognition and classification. Hip dysplasia is the most prevalent hereditary orthopedic disease in the dog. The definitive diagnosis is using the hip radiographic image. This article compares the results of the conventional canine hip dysplasia (CHD) classification by a radiologist using the Fédération Cynologique Internationale criteria and the computer image classification using the Inception-V3, Google’s pre-trained CNN, combined with the transfer learning technique. The experiment’s goal was to measure the accuracy of the model on classifying normal and abnormal images, using a small dataset to train the model. The results were satisfactory considering that, the developed model classified 75% of the analyzed images correctly. However, some improvements are desired and could be achieved in future works by developing a software to select areas of interest from the hip joints and evaluating each hip individually.