DEVELOPMENT OF ATLAS-BASED STRATEGIES FOR THE ANALYSIS OF NEURODEGENERATIVE PATHOLOGIES USING PET/MRI

One of the most interesting future research prospects in the field of PET/MRI imaging is the development of quantitative automated or semi-automated methods for the analysis of neurodegenerative diseases, this projects can be hampered by the lack of a set of functional healthy brain data that can be used to identify pathological anomalies.This thesis focuses on the development of a PET/MRI brain atlas with the aim of obtaining standardizable antomical and functional information. The opportunity for a user to observe a set of neurological imaging data in healthy conditions enables the development of quantitative methods to study the progression of neurodegenerative diseases and to further optimize the diagnostic capabilities and potential treatments, reducing concurrently time loss for the user and the time needed for diagnosis for the patient. The images used in this project were acquired using an integrated full body PET/MRI and subsequently post-processed using SPM in Matlab 2024b environment. SPM was used to execute co-registration, segmentation and normalization operations in order to prepare the PET/MRI images for the atlas creation and while also evaluating time potential co-registration errors related to warping. During post-processing it was also estimated which between PET or MRI would guarantee a better co-registration quality when used as reference image with the aim of reducing potential abnormalities in the brain images. The final atlas is obtained by estimating an average brain anatomical and functional structure thanks to the use of an in-house software and a pool of PET/MRI brain scans. The development of a reliable PET/MRI brain atlas results in the ability of users to recognize anatomical and functional abnormalities in a swifter and simplified matter. Furthermore the practical effect of this development can directly affect ability of a user to analyze more images and consequently reduce time loss needed to deliver a potential personalized treatment to the patient.