Phase unwrapping 4D flow MRI by iterative graph cuts

Motivation: Phase unwrapping 4D flow MRI may alleviate time consuming manual identification of velocity aliasing and the need for re-acquisitions. Goal(s): To apply iterative graph cuts (IGC) to 4D flow MRI. Approach: A performance comparison between 3D and 4D unwrapping was conducted using retrospectively wrapped 4D flow MRI datasets. The performance of the proposed IGC method was also compared to ROMEO and a Laplacian-based algorithm. Results: IGC performance improved with unwrapping of 4D flow compared to 3D flow. IGC performed better than, or comparable to, the reference methods in 3D and 4D unwrapping. Impact: The proposed method shows promise for phase unwrapping 4D flow MRI, extending the dynamic range of 4D flow MRI and reduces the need for multi-VENC acquisitions in clinical cardiac MRI.