Quantifying myelin in crossing fibers using diffusion‐prepared phase imaging: Theory and simulations. Issue 5 (13th July 2021)

• Record Type: Journal Article
• Authors: Cottaar, Michiel; Wu, Wenchuan; Tendler, Benjamin C.; Nagy, Zoltan; Miller, Karla; Jbabdi, Saad
• Is Part Of: Magnetic resonance in medicine. Volume 86:Issue 5(2021)
• Page Start: 2618
• Page End: 2634
• Publication Date: 2021-07-13
• Subjects: diffusion MRI -- magnetic susceptibility -- myelin -- phase imaging -- white matter; Nuclear magnetic resonance -- Periodicals; Electron paramagnetic resonance -- Periodicals; 616.07548
• DOI: 10.1002/mrm.28907
• Languages: English
• ISSNs: 0740-3194

Abstract

Abstract : Purpose: Myelin has long been the target of neuroimaging research. However, most available techniques can only provide a voxel‐averaged estimate of myelin content. In the human brain, white matter fiber pathways connecting different brain areas and carrying different functions often cross each other in the same voxel. A measure that can differentiate the degree of myelination of crossing fibers would provide a more specific marker of myelination. Theory and Methods: One MRI signal property that is sensitive to myelin is the phase accumulation. This sensitivity is used by measuring the phase accumulation of the signal remaining after diffusion‐weighting, which is called diffusion‐prepared phase imaging (DIPPI). Including diffusion‐weighting before estimating the phase accumulation has two distinct advantages for estimating the degree of myelination: (1) It increases the relative contribution of intra‐axonal water, whose phase is related linearly to the thickness of the surrounding myelin (in particular the log g ‐ratio); and (2) it gives directional information, which can be used to distinguish between crossing fibers. Here the DIPPI sequence is described, an approach is proposed to estimate the log g ‐ratio, and simulations are used and DIPPI data acquired in an isotropic phantom to quantify other sources of phase accumulation. Results: The expected bias is estimated in the log g ‐ratio for reasonable in vivo acquisition parameters caused by eddy currents (~4%‐10%), remaining extra‐axonal signal (~15%), and gradients in the bulk off‐resonance field (<10% for most of the brain). Conclusion: This new sequence may provide a g ‐ratio estimate per fiber population crossing within a voxel.