A Hybrid-Two-Compartmental-Model for the Blood-Brain-Barrier Assessment of Glioma Patients

Abstract

Dynamic-contrast-enhanced (DCE)-MRI has shown wide potential in characterizing brain tumor. Conventionally, generalized-tracer-kinetic-model (GTKM) and Patlak-model (PM) have been widely used in clinics to characterize the blood-brain-barrier (BBB) integrity of the tumor. However, PM is recommended for DCE-MRI data with shorter duration (less than 1 minute) and GTKM may overfit in non-enhancing and normal tissues causing erroneous estimation of extra-vascular volume fraction (ve). So, the current study proposes a hybrid tissue depended model with maximum 2 model fitting parameters by combining GTKM and PM depending on underlying tissue. Proposed model has outperformed GTKM in all tissue regions.