Comparative Evaluation of 3D-QALAS Myelin Volume Fraction, Inhomogeneous Magnetization Transfer, and Myelin Water Fraction.
Authors
Affiliations (2)
Affiliations (2)
- From the Department of Radiology (M.N., A.H., N.H., J.K., S.K., X.Z., S.M., C.F., Y.S., S.A., K.K.), Neurology (Y.H., Y.T., T.H.), Juntendo University School of Medicine, Tokyo, Japan; Department of Radiology (M.N., A.H., S.K.), Graduate School of Medicine and Faculty of Medicine, The University of Tokyo, Tokyo, Japan; Center for Medical Imaging Sciences and Visualization (CMIV) (M.W.), Linköping University Hospital, Sweden; Canon Medical Systems Corporation (M.A., H.K., R.Y., M.O.), Tochigi, Japan and Department of Radiology (M.H.), Toho University Omori Medical Center, Tokyo, Japan.
- From the Department of Radiology (M.N., A.H., N.H., J.K., S.K., X.Z., S.M., C.F., Y.S., S.A., K.K.), Neurology (Y.H., Y.T., T.H.), Juntendo University School of Medicine, Tokyo, Japan; Department of Radiology (M.N., A.H., S.K.), Graduate School of Medicine and Faculty of Medicine, The University of Tokyo, Tokyo, Japan; Center for Medical Imaging Sciences and Visualization (CMIV) (M.W.), Linköping University Hospital, Sweden; Canon Medical Systems Corporation (M.A., H.K., R.Y., M.O.), Tochigi, Japan and Department of Radiology (M.H.), Toho University Omori Medical Center, Tokyo, Japan. [email protected].
Abstract
Accurate in vivo quantification of myelin remains challenging despite advances in MRI. We evaluated three-dimensional synthetic MRI-derived myelin volume fraction from three-dimensional quantification using an interleaved Look-Locker acquisition sequence with T2 preparation pulse (3D-QALAS) by comparing it with inhomogeneous magnetization transfer ratio, myelin water fraction, and the T1-weighted/T2-weighted ratio in healthy volunteers and patients with multiple sclerosis. Thirty-one healthy volunteers and 33 consecutive patients with multiple sclerosis underwent 3T MRI including 3D-QALAS, 3D inhomogeneous magnetization transfer imaging, and 2D multi-echo spin-echo imaging for myelin water fraction. Synthetic T1-weighted/T2-weighted ratio maps were generated from quantitative R1, R2, and proton density data. Lesions were segmented using a deep learning-based method, and periplaque regions were defined by 2-voxel isotropic dilation. Atlas-based ROI analyses in normal-appearing brain tissue were performed using repeated-measures correlation. WM lesion-centered analyses were performed to compare metric differences relative to normal-appearing white matter using linear mixed-effects models. In normal-appearing brain tissue pooled across patients and volunteers, myelin volume fraction demonstrated strong overall repeated-measures correlations with the inhomogeneous magnetization transfer ratio (r = 0.86; 95% CI, 0.85-0.87) and the T1-weighted/T2-weighted ratio (r = 0.89; 95% CI, 0.88-0.89), and more modest correlations with myelin water fraction (r = 0.63; 95% CI, 0.61-0.64). The inhomogeneous magnetization transfer ratio-myelin water fraction correlation was r = 0.60 (95% CI, 0.58-0.62). This overall ranking was preserved in volunteers and patients analyzed separately, and Spearman analyses showed a similar pattern. In WM lesion-centered analyses, values relative to normal-appearing WM were lowest for myelin volume fraction (44.2% in plaque, 75.4% in periplaque), followed by the T1-weighted/T2-weighted ratio (56.5%, 77.3%), the inhomogeneous magnetization transfer ratio (74.7%, 87.8%), and myelin water fraction (93.4%, 98.3%). 3D-QALAS-derived myelin volume fraction demonstrated strong overall concordance with the inhomogeneous magnetization transfer ratio in normal-appearing brain tissue and greater lesion-associated deviation from normal-appearing white matter than the other myelin-sensitive metrics. These findings support 3D-QALAS-derived myelin volume fraction as a promising volumetric biomarker of myelin integrity with whole-brain coverage suitable for routine clinical application.