Estimation of contrast agent extravasation from the tissue residue function: application to tumor perfusion Imaging

Purpose: To develop a new method to estimate contrast agent (CA) extravasation directly from the tissue residue function derived from dynamic susceptibility contrast (DSC) MR imaging and to compare the method to a published method for extravasation correction in patients diagnosed with intraaxial brain tumors. Background: CA extravasation is a confounding factor is DSC imaging of brain tumors since it may lead to incorrect estimation of tumor blood volume. Boxerman et al suggested a method to correct for extravasation by assuming CA relaxation to be dominated by T1-shortening in the extravascular, extracellular space (EES) and then correct for leakage relative to an average first-pass response for the entire image slice [1]. This method implicitly assumes that the variation in tissue mean transit times (MTT) is small relative to the slice average MTT. MTT values which deviate significantly from the mean MTT value may, however lead to incorrect estimations of CA extravasation since this may mimic the effect of extravasation in the first-pass response. As an alternative approach, we propose to estimate extravasation directly from the tissue residue function. This approach is insensitive to variations in MTT and further provides a direct extravasation correction of both tissue flow and MTT in addition to CBV. Theory: The estimated tissue CA concentration, Ct(t), including terms for both flow and extravasation can be expressed as: