NMR relaxation times of trabecular bone—reproducibility, relationships to tissue structure and effects of sample freezing

Abstract

Nuclear magnetic resonance (NMR) spectroscopy provides a potential tool for non-invasive evaluation of the trabecular bone structure. The objective of this study was to determine the reproducibility of the NMR relaxation parameters (T2, Carr–Purcel-T2, T1ρ) for fat and water and relate those to the structural parameters obtained by micro-computed tomography (μCT). Especially, we aimed to evaluate the effect of freezing on the relaxation parameters. For storing bone samples, freezing is the standard procedure during which the biochemical and cellular organization of the bone marrow may be affected. Bovine trabecular bone samples were stored at −20 °C for 7 days and measured by NMR spectroscopy before and after freezing. The reproducibility of NMR relaxation parameters, as expressed by the coefficient of variation, ranged from 3.1% to 27.9%. In fresh samples, some correlations between NMR and structural parameters (Tb.N, Tb.Sp) were significant (e.g. the relaxation rate for T2 of fat versus Tb.Sp: r = −0.716, p < 0.01). Freezing did not significantly change the NMR relaxation times but the correlations between relaxation parameters and the μCT structural parameters were not statistically significant after freezing, suggesting some nonsystematic alterations of the marrow structure. Therefore, the use of frozen bone samples for NMR relaxation studies may provide inferior information about the trabecular bone structure.