Abstract

Advisor: Verónica Vilaplana

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

This project attempts to study the cerebral atrophy patterns in gray matter across the
dierent stages of the Alzheimer's Disease (AD), or more specically, along the entire AD
continuum, in a voxelwise approach. To this end, we propose and implement an extensible
toolbox that allows to t dierent models to the data, hence dening a curve for each voxel
that shows the evolution of the gray matter volume in the respective region as compared
to the progression of the disease. The toolbox also includes several evaluation methods
to estimate how closely the proposed model ts the data for each particular voxel. The
resulting values, namely tting-scores, serve as a base to achieve two dierent goals: a)
to identify the regions within the brain that are (most) likely to follow the curve-shape
specied in a given model, and b) to depict the model that best describes the behavior of
the gray matter volume in each voxel from a xed set of models.