Test-retest reproducibility of in vivo magnetization transfer ratio and saturation index in mice at 9.4 Tesla

Abstract

BackgroundMagnetization transfer saturation (MTsat) imaging was developed to reduce T1 dependence and improve specificity to myelin compared to the widely used MT ratio (MTR), while maintaining a feasible scan time. Knowledge of MTsat reproducibility is necessary to apply MTsat in preclinical neuroimaging. PurposeTo assess the test-retest reproducibility of MTR and MTsat in the mouse brain at 9.4 T and calculate sample sizes required to detect various effect sizes. Study TypeProspective Animal ModelC57Bl/6 Mouse Model (6 females and 6 males, aged 12 - 14 weeks) Field Strength/SequenceMagnetization Transfer Imaging at 9.4 T AssessmentAll mice were scanned at two timepoints (5 days apart). MTR and MTsat maps were analyzed using mean region-of-interest (ROI), and whole brain voxel-wise analysis. Statistical TestsBland-Altman plots assessed biases between test and retest measurements. Test-retest reproducibility was evaluated via between and within-subject coefficients of variation (CV). Sample sizes required were calculated (at a 95 % significance level and power of 80 %), given various minimum detectable effect sizes, using both between and within-subject approaches. ResultsBland-Altman plots showed negligible biases between test and retest sessions. ROI-based and voxel-wise CVs revealed high reproducibility for both MTR (ROI: CVs < 8 %) and MTsat (ROI: CVs < 10 %). With a sample size of 6, changes on the order of 15% can be detected in MTR and MTsat, both between and within subjects, while smaller changes (6 - 8 %) require sample sizes of 10 - 15 for MTR, and 15 - 20 for MTsat. Data ConclusionMTsat exhibits comparable reproducibility to MTR, while providing sensitivity to myelin with less T1 dependence than MTR. Our findings suggest both MTR and MTsat can detect moderate changes, common in pathologies, with feasible preclinical sample sizes.