Imaging T1, T2 and myelin water fraction in the post-mortem multiple sclerosis central nervous system

The subject of this thesis is the use of Magnetic Resonance (MR) Imaging to quantify biometric MR indices in the Multiple Sclerosis (MS) fixed post-mortem central nervous system (CNS) tissue. Evaluating these indices in fixed tissue allows for the use of histology to verify the findings of MRI. However, it must first be discovered if the indices can be evaluated in fixed post-mortem spinal cord tissue. There is very little literature in this specific area, though some in the fixed brain, the results of which have been assumed to be equivalent in the spinal cord without proof. Therefore, the methodology must first be verified before the consideration of any index as useful and translatable to in-vivo spinal cord. This thesis concentrates on the evaluation of MR relaxometry methods using the indices T1 and T2 by themselves and to evaluate the myelin content of fixed post-mortem CNS tissue. The Carr-Purcell-Meiboom-Gill (CPMG) and Multicomponent Driven Equilibrium Single Pulse Observation of T1 & T2 (mcDESPOT) sequences are used to calculate T1, T2 and the Myelin Water Fraction (MWF) which is believed to be proportional to myelin content in the CNS. This is performed at 3T in a clinical scanner and at 7T in a small animal and wholebody scanner. The methods are first evaluated for use in fixed post-mortem CNS tissue. The two myelin measurement methods are then compared to histological staining if appropriate and where available to verify that the results obtained are proportional to myelin content. The T1 and T2 values in fixed tissue were found to be shortened in fixed tissue, T2 values were so short as to be at the limits of measurement by a clinical scanner, and values converged in white and grey matter, and therefore contrast was found to be limited between these tissues. Proton density images provided the most contrast between tissues. However, even with shortened T2 values, the CPMG sequence was able to identify the myelin water component in fixed tissue. The mcDESPOT algorithm struggled to separate the myelin water component due to clinical scanner limitations and the shortened, converging T1 and T2 values. However, the mcDESPOT algorithm was successful in discerning the myelin water component in the high signal situation of a small bore 7T peclinical scanner. An evaluation was then made of the usefulness of these indices for translation into clinical imaging. The CPMG sequence was found to be proportional to myelin content under all conditions, and therefore useful for disease monitoring in demyelinating diseases. The mcDESPOT sequence, was found to be proportional to myelin in some conditions, and is likely to be useful for monitoring myelination, though the sequence could not be fully validated in this thesis.