While 2D phase-contrast MRI is often used to examine intracranial vessels in neurovascular disease contexts, the ability of 4D flow to assess many vessels at once makes it an attractive alternative. We aimed to assess the repeatability, reliability, and conformity of 2D and 4D flow across intracranial vessels.
Using correlation analyses and paired
Repeatability for PI measurements was mostly classed as good using both 2D (median ICCβ=β0.765) and 4D (0.772) methods, and for mean flow was mostly moderate across both (2D: 0.711, 4D: 0.571). 4D reliability was good for PI (0.877β0.906) and moderate for mean flow (0.459β0.723). Arterial PI measurements were generally higher using the 2D method, while mean flow was mostly higher using 4D flow.
These results imply that PI measurement using 4D flow is repeatable and reliable across intracranial arteries and veins, but care should be paid to absolute flow measurements as they are susceptible to variation depending on slice placement, resolution, and lumen segmentation practices.