The common approach to myelin mapping relies on multi-T2 component (mcT2) analysis, where the signal from a single voxel is separated into its underlying distribution of T2 values. This approach is highly challenging due to the ill posedness of extracting multiple free parameters from a single voxel data. We present a new data-driven paradigm, where the white matter is first analyzed to identify a finite set of multi-T2 distributions, which are then used to locally analyze the signal in each voxel. Application on white matter tissue produced improved myelin quantifications, without a priory fixing the number of sub-voxel compartments.
This abstract and the presentation materials are available to members only; a login is required.