Abstract
Continuous ambulatory peritoneal dialysis (CAPD) is a treatment used by patients in the end-stage of chronic kidney diseases. Those patients need to be monitored using blood tests and those tests can present some patterns or correlations. It could be meaningful to apply data mining (DM) to the data collected from those tests. To discover patterns from meaningless data, it becomes crucial to use DM techniques. DM is an emerging field that is currently being used in machine learning to train machines to later aid health professionals in their decision-making process. The classification process can found patterns useful to understand the patients' health development and to medically act according to such results. Thus, this study focuses on testing a set of DM algorithms that may help in classifying the values of serum creatinine in patients undergoing CAPD procedures. Therefore, it is intended to classify the values of serum creatinine according to assigned quartiles. The better results obtained were highly satisfactory, reaching accuracy rate values of approximately 95%, and low relative absolute error values.

Abstract
Serious Games apply game strategies to a training environment to encourage participants to make decisions and face challenges - the more interactive, the greater the participants' involvement with the content. Moreover, the best way to train is to simulate and identify scenarios for decision making, recreating situations, and strategies for learning. The Serious Games for training have this purpose. A Serious Game for training can be refined with a game narrative, a methodology centered on group experience defining problems and giving solutions through the game story. The challenge is how to diversify a unique narrative according to the individual player's experience. The present study aims to answer, using Design Science Research, whether a personalized narrative can improve the design of serious games for training. The specific goal is to design, develop and evaluate an artifact based on Design Thinking to design a personalized narrative method for Serious Games.

Abstract

Abstract
Backed by more powerful computational resources and optimized training routines, deep learning models have attained unprecedented performance in extracting information from chest X-ray data. Preceding other tasks, an automated abnormality detection stage can be useful to prioritize certain exams and enable a more efficient clinical workflow. However, the presence of image artifacts such as lettering often generates a harmful bias in the classifier, leading to an increase of false positive results. Consequently, healthcare would benefit from a system that selects the thoracic region of interest prior to deciding whether an image is possibly pathologic. The current work tackles this binary classification exercise using an attention-driven and spatially unsupervised Spatial Transformer Network (STN). The results indicate that the STN achieves similar results to using YOLO-cropped images, with fewer computational expenses and without the need for localization labels. More specifically, the system is able to distinguish between normal and abnormal CheXpert images with a mean AUC of 84.22%.

Abstract
In source localization problems, the relative geometry between sensors and source will influence the localization performance. The optimum configuration of sensors depends on the measurements used for the source location estimation, how these measurements are affected by noise, the positions of the source, and the criteria used to evaluate the localization performance. This paper addresses the problem of optimum sensor placement in a plane for the localization of an underwater vehicle moving in 3D. We consider sets of sensors that measure the distance to the vehicle and model the measurement noises with distance dependent covariances. We develop a genetic algorithm and analyze both single and multi-objective problems. In the former, we consider as the evaluation metric the arithmetic average along the vehicle trajectory of the maximum eigenvalue of the inverse of the Fisher information matrix. In the latter, we estimate the Pareto front of pairs of common criteria based on the Fisher information matrix and analyze the evolution of the sensor positioning for the different criteria. To validate the algorithm, we initially compare results with a case with a known optimal solution and constant measurement covariances, obtaining deviations from the optimal less than 0.1%. Posterior, we present results for an underwater vehicle performing a lawn-mower maneuver and a spiral descent maneuver. We also present results restricting the allowed positions for the sensors.

Abstract
The main goal of this paper is to investigate the effect of different types of self-representations through floating members (hands vs. hands + feet), virtual full body (hands + feet vs. full-body avatar), walking fidelity (static feet, simulated walking, real walking), and number of tracking points used (head + hands, head + hands + feet, head + hands + feet + hip) on the sense of presence and embodiment through questionnaires. The sample consisted of 98 participants divided into a total of six conditions in a between-subjects design. The HTC Vive headset, controllers, and trackers were used to perform the experiment. Users were tasked to find a series of hidden objects in a virtual environment and place them in a travel bag. We concluded that (1) the addition of feet to floating hands can impair the experienced realism (p = 0.039), (2) both floating members and full-body avatars can be used without affecting presence and embodiment (p > 0.05) as long as there is the same level of control over the self-representation, (3) simulated walking scores of presence and embodiment were similar when compared to static feet and real walking tracking data (p > 0.05), and (4) adding hip tracking overhead, hand and feet tracking (when using a full-body avatar) allows for a more realistic response to stimuli (p = 0.002) and a higher overall feeling of embodiment (p = 0.023).