< Back to previous page

Publication

Pattern recognition in neuroimaging for neurodegenerative disorders.

Book - Dissertation

In 2015 , the first simultaneous time-of-flight PET/MR scanner in Belgium (and only the fifth in Europe) was installed in Leuven for clinical use. Besides the huge clinical potential, new developments in multimodal image analysis can provide additional understandings into neurodegenerative disorders that were not clearly feasible with just one modality alone by exploiting the complementary nature of combined PET/MR imaging. In concreto, the objective is to combine structural (SC) and functional (FC) connectivity information derived from multiparametric MR data with the simultaneous acquired metabolic connectivity (MC) information from PET. A graph approach is intended, representing the brain as a network constituted by nodes (parcellated brain regions) and edges expressing the connection strength between the nodes. SC, FC and MC graphs will constructed and metrics describing network topology will be derived. By embedding the brain connectivity graphs in a machine learning framework, the early diagnosis of neurodegenerative disorders will be facilitated. In addition, longitudinal data will be provided, such that a deeper understanding of disease progression will be obtained by investigating how network parameters change over time.
Publication year:2020
Accessibility:Closed