PURPOSE: To improve the multislice performance of velocity-selective arterial spin labeling (VS-ASL) imaging for cerebral blood flow (CBF) measurement such that it might be routinely applied for clinical applications with whole brain coverage. MATERIALS AND METHODS: VS-ASL was performed with improvements such as timing optimization, stimulated echo removal, and slice profile sharpening. Each improvement was evaluated in volunteers by measuring temporal noise in the CBF measurement. VS-ASL with all these improvements was performed in 20 patients with Moyamoya disease some of whom also underwent xenon-enhanced CT (xeCT) imaging which was the reference standard for CBF measurement. RESULTS: Sequence timing optimization and inter-slice crosstalk reduction using stimulated echo removal and slice profile sharpening all contributed to reduction of temporal noise. VS-ASL imaging with all these improvements performed in Moyamoya disease patients showed significant reduction of temporal noise (P < 0.0001) and increased correlation coefficient with xeCT CBF imaging (from 0.07 to 0.62). CONCLUSION: We demonstrated that timing optimization, stimulated echo removal, and slice profile improvement have a large effect on image quality and robustness of VS-ASL in clinical imaging applications.
