Reproducibility of inhomogeneous magnetization transfer (ihMT): A test-retest, multi-site study.

Background: Inhomogeneous magnetization transfer (ihMT) has been reported to feature superior sensitivity and specificity for myelin imaging. However, the reproducibility on ihMT has yet been rarely investigated up to date. The purpose of the present study is to assess the multi-center reproducibility and test-retest variability of ihMT in central nervous system.

Methods: 5 volunteers were recruited and scanned twice on three 3.0 T magnetic resonance imaging (MRI) scanners using ihMT with identical scan parameters. The maps of quantitative ihMT (qihMT) and ihMT ratio (ihMTR) for each scan were calculated. Voxel based analysis then was performed to generate qihMT and ihMTR values for major white matter fibers. The intra- and inter-scanner reliability and reproducibility was assessed with intraclass correlation coefficients (ICCs). Bland-Altman method was used to show the level of agreement between two measurement types. Paired t-test and one-way ANOVA test were also used to compare the difference between inter- and intra-scanner, respectively.

Results: In the 10 major white matter tracts areas, the ICCs indicated high intra- and inter-scanner measurement reliability and reproducibility. The Bland-Altman plots together with 95% confidence interval (CI) across all ROIs in the five volunteers also demonstrated good repeatability. No significant inter- and intra-scanner differences were found in Paired t-test and one-way ANOVA tests.

Conclusion: Good inter- and intra-scanner reliability and reproducibility of ihMT measurements were observed in this study. These findings support the use of ihMT measurements as biomarkers in multicenter and/or longitudinal studies.