The aim of this study is to develop a signal intensity (S(t)) form of the standard Tofts pharmacokinetic model that avoids the need to calculate tissue contrast agent concentration (C(t)) as function of time (t). We refer to this as ‘SI-Tofts’ model. Physiological parameters (Ktrans and ve) calculated using the SI-Tofts and standard Tofts models were compared by using simulations and human prostate dynamic contrast enhanced (DCE) MRI data. This approach was also applied to the Patlak model to compare Ktrans values calculated from C(t) and S(t). Simulations were performed on simulated DCE-MRI data from the Quantitative Imaging Biomarkers Alliance (QIBA) to validate SI-Tofts model. In addition, ultrafast DCE-MRI data were acquired from 18 prostate cancer patients on a Philips Achieva 3T-TX scanner. Regions-of-interest (ROIs) for prostate cancer, normal tissue, gluteal muscle, and iliac artery were manually traced. The C(t) was calculated for each ROI using the standard model with measured pre-contrast tissue T1 values. Both the simulation and clinical results showed strong correlation (r = 0.87 to 0.99, p < 0.001) for Ktrans and ve calculated from the SI-Tofts and standard Tofts models. The SI-Tofts model with a correction factor using the T1 ratio of blood to tissue significantly improved the Ktrans estimates. The correlation of Ktrans obtained from the Patlak model with C(t) vs. S(t) was also strong (r = 0.95 to 0.99, p < 0.001). These preliminary results suggest that physiological parameters from DCE-MRI can be reliably estimated from the SI-Tofts model without contrast agent concentration calculation.
© 2020 Institute of Physics and Engineering in Medicine.

Author