Open Collections

Gradient and spin echo magnetic resonance imaging for the characterization of myelin health in multiple sclerosis

University of British Columbia

Multiple sclerosis (MS) is a complex, autoimmune disease that results in demyelination and neurodegeneration. Magnetic resonance imaging (MRI) is an essential cornerstone of MS diagnosis and clinical decision making, however, clinical MRIs lack specificity to the pathological mechanisms influencing myelin health. In this thesis, two quantitative MRI techniques were probed for their potential to study myelin health in MS. First, multi-echo spin-echo myelin water imaging (MWI) was tested for its specificity to myelin lipids, proteins and iron. We demonstrated to-date unknown sensitivity of MWI and identified MS lesion changes indicative of late-stage remyelination. Thereafter, the accuracy of MWI and its potential application at ultra-high magnetic fields were investigated. Using signal simulations, the dependence of the non-negative least-squares analysis on processing and tissue parameters was described. Myelin underestimations due to B⁺₁-inhomogeneities and noise were shown to be minimized by adjusting the T₂ range according to the echo time. To translate MWI to 7T, T₂ tissue properties in seven healthy subjects were studied in comparison to 3T. We demonstrated the feasibility of 7T-MWI and discussed current limitations in assessing short T₂ times. Secondly, susceptibility-sensitive MRI was explored, which provides greater sensitivity, albeit possibly lower specificity to myelin, than MWI. Using the phase component, we showed that the MS lesion contrast is typically not driven by iron accumulation. In simulations and with post-mortem data, it was demonstrated that iron and myelin loss in combination determine the lesions’ appearance. Thereafter, the potential of the technique to become a marker of tissue damage and repair was evaluated by studying the evolution and pathological underpinnings of acute MS lesions in eleven patients over five years. Current models and their shortcomings were discussed. Finally, two technical developments were introduced. First, a multi-dynamic, high-spatial resolution susceptibility-sensitive imaging approach was presented for visualizing the central vein sign. Using phantom and in vivo data, qualitative and quantitative agreement of the proposed approach with other imaging strategies was demonstrated. Secondly, FLAIR² was introduced, a novel contrast that improves contrast-to-noise, while shortening scan time. The potential of FLAIR² to aid automated lesion segmentation was demonstrated on real-world multi-centre clinical data.

Text

2020-04-30

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/74247

Physics

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/