Room: Room 205
Purpose: To show that the reflux rate (kep) - a promising bio-marker for predicting cancer therapy response - can be estimated through dynamic contrast enhanced (DCE) MRI without requiring contrast agent quantification.
Methods: The Tofts-Kety model (TKM) and the reference region model (RRM) were applied on a virtual phantom from the Quantitative Imaging Biomarkers Alliance, and on clinical DCE-MRI data acquired from 11 patients with glioblastoma multiforme. The models were fitted twice: once by using the signal enhancement ratio, and once by quantifying the tracer concentration. The signal enhancement ratio is defined as (S(t)-S0)/S0 where S(t) is the signal at time t and S0 is the pre-contrast signal. The concentration quantification requires a T1 map which was known for the virtual phantom and was estimated for the clinical cases using variable flip angle data with angles of 2, 5, 10, 15, 20, and 25 degrees.
Results: On the virtual phantom, the RRM was accurate using the signal enhancement ratio, providing kep estimates within 15% of the ground truth, whereas the TKM had errors exceeding 100% on the same data. Both the RRM and TKM were accurate when using the concentration data. On clinical data, both the signal enhancement and tracer concentration data yielded similar kep estimates with the RRM whereas the TKM estimates were dissimilar. The concordance correlation coefficient between the kep estimates using signal and concentration was 0.959 with the RRM and 0.741 with the TKM.
Conclusion: The RRM fits provided accurate kep estimates using either the signal enhancement ratio or the tracer concentration, whereas the TKM was only accurate using the tracer concentration. The benefit of using the signal enhancement ratio is that it is simpler to compute, whereas the concentration quantification requires additional scans combined with error-prone computations.
Funding Support, Disclosures, and Conflict of Interest: Z.A. acknowledges partial support by the CREATE Medical Physics Research Training Network grant of the Natural Sciences and Engineering Research Council (Grant number: 432290).
Pharmacokinetic Modeling, Quantitative Imaging, Perfusion Imaging