Enhancing accuracy and precision of arterial cerebral blood volume imaging in patients with moyamoya disease: Integration of long-labeled long-delay acquisition into DANTE-prepared multi-delay arterial spin labeling.

BACKGROUND: Combining the delays alternating with nutation for tailored excitation (DANTE) pulse with multi-delay arterial spin labeling (MD-ASL) enables arterial cerebral blood volume (CBVa) quantification. However, CBVa estimation may be compromised in moyamoya disease (MMD) because of prolonged arterial transit time (ATT). NEW METHOD: We developed a DANTE-prepared MD-ASL protocol incorporating a long-labeled long-delay (LLLD) acquisition to improve CBVa quantification in the presence of prolonged ATT. RESULTS: Numerical simulations demonstrated that the proposed protocol had smaller negative biases and narrower limits of agreement, indicating improved estimation accuracy and precision. In healthy participants, inclusion of the LLLD acquisition improved test-retest reproducibility. In patients with MMD, the proposed protocol yielded significantly higher CBVa values than the conventional protocol in both short and long ATT regions (short, 0.932 ± 0.218 vs. 0.918 ± 0.211mL/100g, P < 0.001; long, 0.663 ± 0.258 vs. 0.587 ± 0.228mL/100g, P < 0.001). COMPARISON WITH EXISTING METHODS: Compared with the conventional DANTE-prepared MD-ASL protocol, the proposed protocol reduced CBVa underestimation and improved estimation accuracy and precision, particularly in regions with prolonged ATT. CONCLUSIONS: The LLLD acquisition mitigates CBVa underestimation associated with prolonged ATT. Integrating an LLLD acquisition into DANTE-prepared MD-ASL improves the accuracy and precision of CBVa estimation, enabling more reliable noninvasive assessment of cerebral hemodynamics in patients with MMD.