Optimizing myelin water imaging in multiple sclerosis

Myelin water imaging (MWI) allows in vivo quantification of myelin content using MRI. Multi-echo gradient-echo (MEGRE) sequences with Variable Flip Angles (VFA), combined with diffusion-weighted imaging (DWI), can generate robust, whole-brain myelin water fraction (MWF) maps more efficiently than conventional implementations. We evaluated how spatial resolution and number of flip angles affect MWI quality in a clinical research setting. Fifteen people with MS and ten healthy controls were scanned at 3 T (Siemens VIDA). MRI included multi-shell DWI and 3D-MEGRE acquired at 2.0 mm and 2.5 mm isotropic resolution. Twelve echoes were acquired at five flip angles. MWF was quantified based on a three-compartment T1, T2*, and DWI model, and determined in normal appearing white matter (NAWM), lesions and corpus callosum. Associations between MWF across resolutions and VFA schemes were assessed using Bland-Altman, ICC-C and ICC-AA. MWF was highest in corpus callosum (10.2 ± 1.1%), intermediate in NAWM (7.9 ± 0.7%), and lowest in lesions (4.9 ± 0.7%) (values for 2.0 mm and FA5). A FA4 MEGRE protocol showed comparable MWF values, while FA3 systematically overestimated MWF and showed greater variability, particularly in lesions. ICC-C between spatial resolutions was good to excellent (0.80-0.91), ICC-AA was moderate to good (0.66-0.81), indicating that the two resolutions should not be treated interchangeably. QC monitoring of fitting results is recommended to strengthen reliability of MWF results. MWI duration can be shortened with a FA4 MEGRE protocol. Resolution choice should be guided by study aims, as 2.0 mm and 2.5 mm were not fully interchangeable.