Abstract
Advisors: Verónica Vilaplana, Adrià Casamitjana
Studies: Bachelor degree in Science and Telecommunication Technologies Engineering at Telecom BCN-ETSETB from the Technical University of Catalonia (UPC)
Alzheimer's disease is still an incurable disease. Nevertheless, some of its biomarkers suffer changes in the early stages of the disease, long before clinical symptoms appear. In order to determine how biomarkers obtained from magnetic resonance (MRI) techniques affect the disease's evolution, machine learning techniques have been used to design and implement a classification system so as to predict the stages in which several patients belong. One of the main objectives of this project is reducing the number of data to manage, since MRI provide a large volume of data for each patient. As a result, we will focus on the stage of reduction and extraction of characteristics of the classifier which may be relevant for the mentioned problem. We will carry out an exhaustive analysis of different methods of selection of features to apply to biomedical data related to Alzheimer's disease. Results obtained will also be applicable to other fields. Finally, we will assess these methods with a multimodal data base provided by the collaboration agreement with Pasqual Maragall Foundation (FPM).