Abstract

Advisor: Verónica Vilaplana

Studies: Bachelor degree in Science and Telecommunication Technologies Engineering at Telecom BCN-ETSETB from the Technical University of Catalonia (UPC)

Neurodegenerative diseases impose substantial public health burdens on populations
throughout the world. Alzheimer's disease is among the major neurodegenerative diseases,
and its causes and treatment are still unknown. Researchers around the world are conducting
large data-driven studies in order to unveil the causes and biological mechanisms of such
diseases, and for that reason automatic tools that allow to uncover statistically signicant
ndings are needed.
To address this problem we present in this thesis a software toolbox that provides the
tools to analyze the linear and nonlinear dynamics of gray-matter and study the statistical
signicance of such dynamics at the voxel level. The toolbox features various tting methods
and t evaluation metrics, an automatic hyperparameters look-up algorithm and several
visualization and comparison tools.
All the features provided in this toolbox were tested in two real problems provided by
the Pasqual Maragall Foundation, and it yielded results that were validated by the ndings
in the original studies.