Multi‐center evaluation of stability and reproducibility of quantitative MRI measures in healthy calf muscles

The purpose of this study was to evaluate temporal stability, multi-center reproducibility and the influence of covariates on a multimodal MR protocol for quantitative muscle imaging and to facilitate its use as a standardized protocol for evaluation of pathology in skeletal muscle. Quantitative T2, quantitative diffusion and four-point Dixon acquisitions of the calf muscles of both legs were repeated within one hour. Sixty-five healthy volunteers (31 females) were included in one of eight 3-T MR systems. Five traveling subjects were examined in six MR scanners. Average values over all slices of water-T2 relaxation time, proton density fat fraction (PDFF) and diffusion metrics were determined for seven muscles. Temporal stability was tested with repeated measured ANOVA and two-way random intraclass correlation coefficient (ICC). Multi-center reproducibility of traveling volunteers was assessed by a two-way mixed ICC. The factors age, body mass index, gender and muscle were tested for covariance. ICCs of temporal stability were between 0.963 and 0.999 for all parameters. Water-T2 relaxation decreased significantly (P < 10^-3 ) within one hour by ~ 1 ms. Multi-center reproducibility showed ICCs within 0.879-0.917 with the lowest ICC for mean diffusivity. Different muscles showed the highest covariance, explaining 20-40% of variance for observed parameters. Standardized acquisition and processing of quantitative muscle MRI data resulted in high comparability among centers. The imaging protocol exhibited high temporal stability over one hour except for water T2 relaxation times. These results show that data pooling is feasible and enables assembling data from patients with neuromuscular diseases, paving the way towards larger studies of rare muscle disorders.