Use this URL to cite or link to this record in EThOS:
Title: Applications of MRI magnetic susceptibility mapping in PET-MRI brain studies
Author: Dixon, Emma Charlotte
ISNI:       0000 0004 7429 147X
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Magnetic susceptibility mapping (SM) uses magnetic resonance imaging (MRI) phase images to produce maps of the magnetic susceptibility (χ) of tissues. This work focuses on the applications of SM-based imaging to PET-MRI, the hybrid imaging modality which combines positron emission tomography (PET) with MRI. First, the potential of using SM to aid PET attenuation correction (AC) is explored. AC for PET-MRI is challenging as PET-MRI provides no information regarding the electron density of tissues. Recently proposed SM methods for calculating the χ in regions of no MRI signal are used to segment air, bone and soft tissue in order to create AC maps. In the head, SM methods are found to produce inferior air/bone segmentations to high-performing AC methods, but result in more accurate AC than ultrashort-echo (UTE)-based air/bone segmentations, and may be able to provide additional information in subjects with atypical anatomy. Secondly, a SM pipeline for inclusion in a PET-MRI study into biomarkers for Alzheimer’s disease (AD) is developed. In the Insight46 study 500 healthy subjects from the 1946 MRC National Survey of Health and Development are undergoing a comprehensive PET-MRI protocol at two time-points. SM processing methods are compared and optimised, and a method for processing images with oblique imaging planes is developed. The effect of using different tools for automated segmentation of regions of interest (ROIs) on reported regional χ values is analysed. The ROIs resulting from different tools are found to result in large differences in χ values. FIRST is chosen as the most appropriate ROI segmentation tool for this study based on anatomical accuracy as assessed by a neuroradiologist. Initial analysis of χ values from 100 subjects using data from the first time-point is carried out. No significant association with regional χ values is found for amyloid status, PET radiotracer uptake, or APOE genotype.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available