Abstract
AbstractInclusions of magnetic materials within brain tissue affect the decay of the MRI signal due to transverse relaxation. Theoretical studies of the gradient-echo signal anticipate a transition of the signal decay from Gaussian at short echo times to exponential at long echo times. However, transverse relaxation signal decay is widely reported to follow a simple exponential behaviour, with a rate. Here, we provide experimental evidence of non-exponential signal decay due to transverse relaxation in human subcortical grey matter, from gradient-echo MRI data acquiredin vivoat 3T.The signal decay follows the qualitative behaviour predicted by theoretical models of the effect of magnetic inclusions on the MRI signal. In subcortical grey matter, such models lead to an improved fit to experimental data than the traditionally used exponential model. Amongst subcortical regions, the strongest deviations from exponential behaviour take place in the substantia nigra and globus pallidus.From the estimates of the signal model parameters, we assess the properties of the magnetic inclusions at the source of the effect within brain tissue. In the limiting case of the static dephasing regime, the magnetic susceptibility and volume fractions of the inclusions range from 1.2 to 2.6 ppm SI units and from 0.03 to 0.06 between subcortical regions, respectively. The higher estimates in the substantia nigra and globus pallidus are consistent withex vivohistological studies and suggest a dominant contribution of iron-rich inclusions. In the limiting case of the diffusion narrowing regime, the typical size of the magnetic inclusions is ∼3 μm, representative of typical iron-containing cells.Non-exponential transverse relaxation signal decay in subcortical grey matter provides new means to characterize the spatial distribution of magnetic inclusions within brain tissue with increased specificity.
Publisher
Cold Spring Harbor Laboratory