Multicenter mapping in the healthy brain

Abstract

PURPOSE: The R2* relaxation rate constant has been suggested as a sensitive measure for iron accumulation. The aim of this multi-center study was to assess the inter-scanner and inter-subject variability of R2* mapping and to investigate the relationship between brain volume and R2* in specific structures. METHODS: R2* mapping was performed in 81 healthy subjects in seven centers using different 3 T systems. R2* was calculated from a dual-echo gradient echo sequence and was assessed in several deep gray matter structures. The inter-scanner and inter-subject variability of R2* was calculated by means of the coefficient of variation before and after correcting for age. RESULTS: Significant center effects were seen in some regions which get lost after age correction. The coefficient of variation for the inter-center variability was much lower (<5.6%) than for the intra-subject variability (6.7%-11.7%). R2* in the putamen and red nucleus scaled with cortical volume while R2* in the globus pallidus and the substantia nigra was negatively associated with white matter volume. CONCLUSION: R2* is a robust and reproducible measure in a multicenter setting provided that a standardized MRI protocol is used. The relationship between iron concentration in deep gray matter and volume of specific brain compartments needs further investigation.