ABSTRACT Purpose We aim to assess the reliability of cerebral blood flow (CBF) and cerebrovascular reactivity (CVR) obtained simultaneously from a novel multi‐band pseudo‐continuous arterial spin labeling (M2‐PCASL) sequence and hypercapnic gas challenge, to examine factors influencing CBF variability, and to evaluate the impact of carbon dioxide (CO 2 ) concentration on CVR estimates. Methods Nine participants underwent test–retest M2‐PCASL scans at 3T with hypercapnia to quantify CBF and CVR. Reliability was assessed using the intraclass correlation coefficient (ICC) at whole brain, region of interest (ROI), and voxel levels. Linear mixed‐effect models were used to investigate factors contributing to CBF variability. Independent samples t ‐tests were used to compare BOLD CVR and temporal contrast to noise ratio (tCNR) from different CO 2 concentrations. Results With smoothing, the M2‐PCASL sequence estimated baseline CBF and BOLD CVR with good to excellent reliability (ICC > 0.81) in the whole brain, gray matter (GM), and white matter (WM). Average ICCs across the Automated Anatomical Labeling atlas ROIs were between 0.50 ± 0.30 and 0.69 ± 0.19. CBF CVR achieved fair to moderate reliability in GM (ICCs between 0.42 and 0.55). Head motion was significantly associated with CBF temporal SNR ( p ≤ 0.002). 5% and 8% CO 2 yielded similar BOLD CVR estimates, but the BOLD tCNR using 5% CO 2 was lower, although not statistically significant. Conclusion The M2‐PCASL sequence with hypercapnia reliably estimates baseline CBF and BOLD CVR at improved spatial and temporal resolutions relative to existing methods, enabling us to noninvasively and comprehensively assess cerebrovascular health.
Todd et al. (Wed,) studied this question.