Motivation: Quantitative neuroimaging studies often require large datasets, and it is common to collect the data using multiple scanners. Goal(s): However, the inter-site reproducibility of DTI parameters should be determined before embarking on a multi-centre study. Approach: We used TractFinder to assess the reproducibility of DTI parameters in clinically relevant WM tracts across three scanners of the same type (Siemens Prisma 3T). Results: Our results provide evidence that FA and MD have excellent reproducibility across scanners of the same type when the acquisition scheme and pre-processing pipeline are identical, and WM tract ROIs are automatically segmented. Impact: DTI-derived FA and MD have excellent reproducibility (ICC ≥ 0.95) across three scanners of the same type when WM tract ROIs are automatically segmented. This allows neuroimaging studies to use multi-centre DTI data with confidence.
Sierhej et al. (Tue,) studied this question.