Metodologías de procesamiento de datos en el ámbito de e-health para la categorización de respuestas terapéuticas en pacientes con migraña

Résumé

This Ph.D. Thesis studies some data processing methodologies in the area of e-Health for categorizing therapeutic responses in patients with migraine. In a real e-Health scenario, this work focuses on the prediction of the response to the treatment of migraine through the use of retrospective medical records collected from Hospital Clínico Universitario in Valladolid and Hospital Universitario de La Princesa, in Madrid. The goal of this research work is to pose and answer the following questions: is it possible to predict the response to every stage of the BoNT-A treatment for migraine? Does a pre-treatment prediction model for the BoNT-A treatment in migraine exist? how do these models respond under missing values? Is it possible to reveal those medical factors that make it possible a high response to the BoNT-A treatment? Are the medical factors used to predict the response of the treatment coherent with the knowledge of medical experts? To answer these questions, this work has explored and implemented different approaches for the training of the predictive models...