Abstract
PURPOSE: To quantify T(2) relaxation in the brain at 3 T and 7 T to study its field dependence and correlation with iron content, and to investigate whether iron can be separated from other sources of T(2) relaxation based on this field dependence. METHODS: Nine subjects were scanned at both field strengths with the same acquisition technique, which used multiple gradient-echo sampling of a spin echo. This allowed for separation of T(2) relaxation from static dephasing by B(0) field inhomogeneities and the effects of radiofrequency refocusing imperfections. The average relaxation rates (R(2) = 1/T(2)) in multiple regions of interest in the brain were fitted with a model linear in B(0) and correlated with literature iron values. RESULTS: The relationship between the R(2) values at the two field strengths appeared to be linear over all regions of interest. The R(2) values (in s(-1)) in the regions of interest for which both an iron and a lipid mass fraction have been documented in the literature were fitted as R2 = 9 + (0.9 + 2 · 104[Fe] + 5.7[lipid]) · B0 , where [Fe] and [lipid] indicate the putative mass fractions of iron and lipid. CONCLUSION: The R(2) relaxation rate is well described by a constant plus a term linear in B(0), with both iron and lipid content contributing to the slope. This indicates that the contributions of lipid and iron to R(2) cannot be separated based solely on the field dependence of R(2) in the field range of 3-7 T.