Relationships Between Young Modulus of Elasticity, Ash Density, and MRI Derived Effective Transverse Relaxation T2* in Tibial Specimens

OBJECTIVE: It has been hypothesized that the MR relaxation time T2* of bone marrow present in the intertrabecular spaces may be related to the density of the trabecular network and may be a predictor of trabecular bone properties. MATERIALS AND METHODS: To derive a relationship between the marrow relaxation time T2* and biomechanical properties of trabecular bone, we studied two sets of trabecular bone specimens from human tibiae. The first set consisted of 12 specimens that were defatted and immersed in saline; the second set consisted of 18 specimens with marrow in the trabecular spaces. The MR studies were conducted on a 1.5 T imaging system. In the first set of specimens, a GE sequence (TR = 70 ms; TE = 5, 10, 15, 20, 25 ms) was used to obtain images in the axial plane. In the second set, a water suppression pulse was used prior to an asymmetric SE sequence (TR = 300 ms; TE = 4, 8, 12, 16, 20, 24 ms) to obtain images in the axial, coronal, and sagittal planes. The T2* of the intertrabecular saline of the marrow fat was calculated assuming a monoexponential decay. In both sets, the elastic moduli were measured in three orthogonal directions (superoinferior, anteroposterior, and mediolateral). The ash density was determined after the completion of the experiments. RESULTS: Our results indicate a moderate significant negative correlation between T2* and ash density or elastic modulus (E) in both sets of specimens. The correlation coefficients were slightly improved between the transverse relaxation rate 1/T2* and bone density or E. We found a good correlation between T2* and the reciprocal ash density (r = 0.88) and between T2* and the reciprocal elastic modulus 1/E (r = 0.87 to r = 0.95) in the first set, while in the second set the correlation remained moderate. With use of a multiple linear regression model (1/E = a x T2* + b x 1/T2* + n), the reciprocal elastic moduli 1/E were predicted to > 90% by T2* and 1/T2* in the first set of specimens. This finding was not replicated with the second set of specimens. In the second set of specimens, we found poor to moderate correlation coefficients between the T2* times in the three orthogonal planes (r = 0.45 to r = 0.71). CONCLUSION: Trabecular bone properties such as density and strength may potentially be assessed with quantitative MR techniques. However, especially for in vitro studies, specimen preparation, acquisition parameters, and specimen geometry may have a significant impact on the obtained results.