T₁ relaxation of bound and pore water in cortical bone

MRI measures of bound and/or pore water concentration in cortical bone offer potential diagnostics of bone fracture risk. The transverse relaxation characteristics of both bound and pore water are relatively well understood and have been used to design clinical MRI pulse sequences to image each water pool quantitatively. However, these methods are also sensitive to longitudinal relaxation characteristics, which have been less well studied. Here, spectroscopic relaxometry measurements of 31 human cortical bone specimens provided a more detailed picture of $T_1$ of both bound and pore water. The results included mean, standard deviation, and range of $T_1$ spectra from both bound and pore water, as well as novel presentations of the 2D $T_1-T_2$ distribution of pore water. Importantly, for each sample the pore water $T_1$ spectrum was found to span more than one order of magnitude and varied substantially across the 31 sample studies. Because many existing methods assume pore water $T_1$ to be mono-exponential and constant across individuals, the results were used to compute the potential effect neglecting this intra- and intersample $T_1$ variation on accurate MRI measurement of both bound and pore water concentrations. The greatest effect was found for adiabatic inversion recovery (AIR) based measurements of bound water concentration, which showed an average of 8.8% and as much as 37% error when using a common mono-exponential assumption of pore water $T_1$ . Despite these errors, the simulated AIR measurements were still moderately well correlated with the bound water concentrations derived from the spectroscopic data.