Robust Multi-TE ASL-Based Blood–Brain Barrier Integrity Measurements

Abstract

Multiple echo-time arterial spin labelling (multi-TE ASL) offers estimation of blood–tissue exchange dynamics by probing the T2 relaxation of the labelled spins. In this study, we provide a recipe for robust assessment of exchange time (Texch) as a proxy measure of blood–brain barrier (BBB) integrity based on a test-retest analysis. This includes a novel scan protocol and an extension of the two-compartment model with an “intra-voxel transit time” (ITT) to address tissue transit effects. With the extended model, we intend to separate the underlying two distinct mechanisms of tissue transit and exchange. The performance of the extended model in comparison with the two-compartment model was evaluated in simulations. Multi-TE ASL sequence with two different bolus durations was used to acquire in vivo data (n = 10). Cerebral blood flow (CBF), arterial transit time (ATT) and Texch were fitted with the two models, and mean grey matter values were compared. Additionally, the extended model also extracted ITT parameter. The test-retest reliability of Texch was assessed for intra-session, inter-session and inter-visit pairs of measurements. Intra-class correlation coefficient (ICC) and within-subject coefficient of variance (CoV) for grey matter were computed to assess the precision of the method. Mean grey matter Texch and ITT values were found to be 227.9 ± 37.9 ms and 310.3 ± 52.9 ms, respectively. Texch estimated by the extended model was 32.6 ± 5.9% lower than the two-compartment model. A significant ICC was observed for all three measures of Texch reliability (P < 0.05). Texch intra-session CoV, inter-session CoV and inter-visit CoV were found to be 6.6%, 7.9%, and 8.4%, respectively. With the described improvements addressing intra-voxel transit effects, multi-TE ASL shows good reproducibility as a non-invasive measure of BBB permeability. These findings offer an encouraging step forward to apply this potential BBB permeability biomarker in clinical research.