Optimizing the vascular signal suppression in magnetic resonance imaging (MRI)

Abstract Hemodynamic parameters offer a rich resource of physiological biomarkers for studying natural aging and various diseases. Blood signals are often suppressed or isolated according to the study purposes to focus on a certain aspect of hemodynamic properties. In the arterial spin labeling (ASL) MRI, vascular hyperintensities introduce significant quantitative bias to the regional comparisons. Velocity encoding module comprising a pair of bipolar gradients with equal first momenta is popularly applied to suppress these vascular signals. However, there is a lack of systematic investigation on how experimental conditions affect vascular signal suppression. Systematic studies involving a series of optimizations covering broad parametric ranges at fine resolutions will lead to prohibitively long scan time. Therefore, as an alternative, our current study utilizes simulations to explore the dependence of vascular suppression on the experimental conditions. We found that vascular suppression with >95% efficiency (i.e., <5% residual signals) could be achieved only when voxel sizes were sufficiently large to cover the whole vessel. The ratio between peak velocity and encoding velocity should be ≥2.976 in order to achieve ≥90% vascular signal suppression. If the flow pattern is a mixture of plug and laminar flows, ≥90% vascular signal suppression can be obtained when laminar flow occupies at least 50% of the total flow.