A method for reducing the effects of motion contamination in arterial spin labeling magnetic resonance imaging

Arterial spin labeling (ASL) is a noninvasive method to measure cerebral blood flow (CBF). Arterial spin labeling is susceptible to artifact generated by head motion; this artifact is propagated through the subtraction procedure required to calculate CBF. We introduce a novel strategy for mitigating this artifact based on weighting tag/control volumes according to a noise estimate. We evaluated this strategy (DVARS weighting) in application to both pulsed ASL (PASL) and pseudo-continuous ASL (pCASL) in a cohort of normal adults (N=57). Application of DVARS weighting significantly improved test-retest repeatability as assessed by the intra-class correlation coefficient. Before the application of DVARS weighting, mean gray matter intra-class correlation (ICC) between subsequent ASL runs was 0.48 and 0.51 in PASL and pCASL, respectively. With weighting, ICC was significantly improved to 0.63 and 0.58.