Quantitative susceptibility mapping and R1 measurement: Determination of the myelin volume fraction in the aging ex vivo rat corpus callosum

Abstract

In studies of the white matter (WM) in aging brains, both quantitative susceptibility mapping (QSM) and direct R-1 measurement offer potentially useful ex vivo MRI tools that allow volumetric characterization of myelin content changes. Despite the technical importance of such MRI methods in numerous age-related diseases, the supposed linear relationship between the estimates of either the QSM or R-1 method and age-affected myelin contents has not been validated. In this study, the absolute myelin volume fraction (MVF) was determined by transmission electron microscopy (TEM) as a gold standard measure for comparison with the values obtained by the aforementioned MR methods. To theoretically evaluate and understand the MR signal characteristics, QSM simulations were performed using the finite perturber method (FPM). Specifically, the simulation geometry modeling was based on TEM-derived structures aligned orthogonally to the main magnetic field, the construct of which was used to estimate the magnetic field shift (Delta B) changes arising from the conjectured myelin structures. Experimentally, ex vivo corpus callosum (CC) samples from rat brains obtained at 6 weeks (n = 3), 4 months (n = 3), and 20 months (n = 3) after birth were used to establish the relationship between changes quantified by either QSM or R-1 with the absolute MVF by TEM. From the ex vivo brain samples, the scatterplot of mean MVF versus R-1 was fitted to a linear equation, where R-1mean = 0.7948 x MVFmean + 0.8118 (Pearson&apos;s correlation coefficient r = 0.9138; p < 0.01), while the scatterplot of mean MVF versus MRI-derived magnetic susceptibility (chi) was also fitted to a line where chi(measured,mean) = -0.1218 x MVFmean - 0.006345 (r = -0.8435; p < 0.01). As a result of the FPM-based QSM simulations, a linearly proportional relationship between the simulated magnetic susceptibility, chi(simulated,mean), and MVF (r = -0.9648; p < 0.01) was established. Such a statistically significant linear correlation between MRI-derived values by the QSM (or R-1) method and MVF demonstrated that variable myelin contents in the WM (i.e., CC) can be quantified across multiple stages of aging. These findings further support that both techniques based on QSM and R-1 provide an efficient means of studying the brain-aging process with accurate volumetric quantification of the myelin content in WM.