Evaluation of Blood-Brain Barrier Integrity by the Analysis of Dynamic Contrast-Enhanced MRI – a Comparison of Quantitative and Semi-Quantitative Methods

Abstract

Disruption of the blood-brain barrier (BBB) is a key feature of various brain disorders. To assess its integrity a parametrization of dynamic magnetic resonance imaging (DCE MRI) with a contrast agent (CA) is broadly used. Parametrization can be done quantitatively or semi-quantitatively. Quantitative methods directly describe BBB permeability but exhibit several drawbacks such as high computation demands, reproducibility issues, or low robustness. Semi-quantitative methods are fast to compute, simply mathematically described, and robust, however, they do not describe the status of BBB directly but only as a variation of CA concentration in measured tissue. Our goal was to elucidate differences between five semi-quantitative parameters: maximal intensity (Imax), normalized permeability index (NPI), and difference in DCE values between three timepoints: baseline, 5 min, and 15 min (Δ5-0, Δ15-0, Δ15-5) and two quantitative parameters: transfer constant (Ktrans) and an extravascular fraction (Ve). For the purpose of comparison, we analyzed DCE data of four patients 12-15 days after the stroke with visible CA enhancement. Calculated parameters showed abnormalities spatially corresponding with the ischemic lesion, however, findings in individual parameters morphometrically differed. Ktrans and Ve were highly correlated. Δ5-0 and Δ15-0 were prominent in regions with rapid CA enhancement and highly correlated with Ktrans. Abnormalities in Δ15-5 and NPI were more homogenous with less variable values, smoother borders, and less detail than Ktrans. Moreover, only Δ15-5 and NPI were able to distinguish vessels from extravascular space. Our comparison provides important knowledge for understanding and interpreting parameters derived from DCE MRI by both quantitative and semi-quantitative methods.