Quantitative magnetic resonance imaging of magnetization transfer and T2 relaxation in human white matter pathology

The primary aim of this thesis is the reconciliation of two seemingly disparate quantitative magnetic resonance imaging (MRI) techniques proposed to characterize human brain white matter (WM) in health and disease. Quantitative magnetization transfer imaging (QMTI) and multi-component analysis of T2 relaxation (QT2) both attempt to quantify myelin content in vivo, but are based on fundamentally different models of WM. QMTI probes the macromolecular component of tissue using a two-pool model of magnetization transfer, while QT2 isolates the water signal from distinct micro-anatomical compartments. The specific objectives were to determine the interrelationship between measurements made with both techniques in the context of potential pathological changes associated with multiple sclerosis (MS), and to apply both to track WM changes in the acute phase of MS lesions. First, simulations were used to evaluate the theoretical sensitivity of each technique to the characteristics of a model of WM that incorporates four pools of magnetization, based on published in vitro measurements. Next, the experimental reproducibility of each technique was investigated, and the impact of certain basic variations in the data acquisition and analysis procedures was evaluated. In the final stage, both methods were applied longitudinally in vivo to assess the dynamic changes that occur in acute, contrast-enhancing lesions of MS. The theoretical results illustrate the sensitivity and limitations of QMTI and QT2 to specific pathology-inspired modifications of WM, and shed new light on the potential specificity of often-neglected QMTI parameters. The reproducibility of both techniques is acceptable for use in repeated clinical measurements, and QMTI has lower variability overall. The importance of corrections for magnetic field inhomogeneity in QMTI is demonstrated, and a simple optimization of the QMTI data acquisition is introduced. Both techniques were sensitive to active disease pathology in the longitudinal study of MS patients. Overall, this thesis demonstrates the complementary nature and usefulness of QMTI and QT2 in the characterization of the natural disease course of a degenerative disease of the human central nervous system. With further refinement, these techniques could play an important role in the study of other diseases, and have the potential to serve as outcome measures in clinical trials.