Anne van den Brekel, Marnix C Maas, Ilse M Spenkelink, Willem C M van Tilburg, Joey Roosen, Daniel Giese, Ning Jin, Jurgen J Fütterer, J Frank W Nijsen
Objective: To investigate the feasibility and repeatability of four-dimensional (4D) flow MRI for quantifying hepatic arterial blood flow (HABF) in healthy volunteers and patients with liver disease, and to identify sources of measurement variability.
Materials and methods: In this prospective single-centre study, 23 healthy volunteers and 12 patients underwent two-dimensional (2D) and 4D flow acquisitions at 3T. Time-averaged and peak systolic vessel areas, velocities, and flow rates of the common hepatic artery (CHA) were extracted. Feasibility of 4D flow was assessed via CHA visibility on phase-contrast MRA. Scan-rescan repeatability of 4D flow was assessed by calculating within-subject coefficient of variation (CVwithin) using two-level variance components analysis (VCA). Within-subject variance was further decomposed into acquisition-related and plane positioning components using three-level VCA. Agreement between 2D and 4D flow was evaluated using Bland-Altman analysis and Pearson's correlation.
Results: Average 4D flow feasibility rate was 83%. CVwithin ranged from 12-30%, with acquisition-related factors and plane positioning accounting for 30-40% and 7-11%, respectively. 4D flow overestimated vessel area by ~ 18% and underestimated velocity and flow rate by ~ 53% and ~ 36%, respectively, compared to 2D flow.
Discussion: 4D flow MRI enables HABF quantification, though substantial scan-rescan variability exists. Acquisition-related factors were the main source of within-subject variability, while plane positioning contributed less.