Measurement of the concentration of the contrast agent using dynamic susceptibility contrast MRI relies on field inhomogeneities caused by the presence of the paramagnetic agent. The usual method for calculation of the concentration from dynamic T2*-weighted images is based on two key assumptions: 1) a linear relation between the change in R2* and the contrast agent concentration, and 2) a negligible effect on the MR signal due to concurrent T1 changes. In this study the effect of inaccuracies in these two assumptions on perfusion measurements was investigated using simulations and in vivo data. The results of the simulations provide a quantitative characterization of the magnitude of these effects for various experimental conditions (e.g., when a 1-sec TR is used with TE=20 ms, the T1 effects can introduce up to 40% cerebral blood flow underestimation depending on the flip angle). These findings can be used as a guide to estimate the errors in specific practical implementations, as well as to optimize the sequence parameters to minimize their effect. In summary, this study shows that the arterial input function measurement should be corrected for nonlinear R2* effects and that care should be taken in the study design to avoid introducing significant T1 effects in perfusion quantification.
Copyright (c) 2007 Wiley-Liss, Inc.