Multi-PLD ASL avoids up to 30% CBF underestimation in above half subjects without compromising test-retest reliability compared to single-PLD

Motivation: Long arterial transit time (ATT) may cause underestimation of CBF in single-delay arterial spin labeling (ASL) quantification.Goal(s): Our goal is to evaluate the accuracy and reliability between single- and multi-delay ASL acquisition. Approach: 28 subjects underwent test-retest scans ~1 week apart. Voxel-wise and regional CBF/ATT values were quantified to evaluate the test-retest reliability. ATT values obtained from 5-PLD data were used to estimate quantification errors for CBF estimated from single-delay (PLD=2s) ASL. Results: Overall, single-delay may cause an average of 4% to 30% of CBF underestimation in 56% regions. 5-delay is a solid solution to evaluate ATT and CBF. Impact: While single-delay ASL results in slightly higher test-retest reliability, the underestimation of CBF may compromise the quantification accuracy. Since ATT variation is more common in elderly and patients, multi-delay ASL with model-fitting analysis is expected to outperform single-delay ASL.