Abstract
Maguetic field inhomogeneities due to differences in susceptibility between trabecular bone and bone marrow result in a reduction in T2*. The authors previously quantified the relationship between the relaxation rate enhancement per unit change in bone density, Δ*, using dried, excised vertebral bodies immersed in saline. In the present study, they investigated the precision and reproducibility of such measurements in vitro and found that the short‐term precision ranges from 2% to 11%, while the longterm precision error, which may be governed by the placement of the region of interest, can vary up to 50%. A chemical shift—selective asymmetric spin‐echo sequence was used to assess T2* changes in the saturated fat component of bone marrow in vivo. It was shown that the marrow fat relaxation rate increases as the surrounding trabecular bone density increases and that the Δ* of the marrow fat component was 0.20 sec −1/mg/cm3. The results also indicate that the distribution of T2* varies with image resolution. Both in vitro and in vivo, characteristics of the relaxation time distribution such as the mean, standard deviation, and skewness decrease as image resolution decreases, the degree of variation depending on the density of the surrounding trabecular network.