Adaptive Respiratory Motion Correction Framework for Ungated Hepatic 4D Flow MRI
Ruiyu Cao1, Shaohang Li1, Aiqi Sun2
1Institute of Science and Technology for Brain-Inspired Intelligence, Fudan University, Shanghai, China.
Purpose:
To suppress respiratory motion-induced artifacts in ungated hepatic 4D flow MRI, improving the accuracy of flow quantification and small-vessel visualization while maintaining predictable scan times.
Methods:
An adaptive respiratory motion correction framework (AMOCO) was developed, integrating optimized sampling, adaptive data selection, and combined intra-bin and inter-bin motion correction. The method was evaluated in 15 healthy volunteers and one patient with liver cirrhosis on a 5 T system, and further validated in 5 healthy volunteers on a 3 T system. Ungated liver 4D flow reconstructed with AMOCO was compared with 2D flow and navigator-gated 4D flow references. Image quality was assessed using gradient entropy, and flow measurements in the portal vein and its left and right branches were analyzed. Statistical group differences, correlations, and measurement consistency were evaluated. The feasibility of the AMOCO method applied across 5T and 3T scanning platforms was preliminarily validated.
Results:
Ungated 4D flow MRI achieved clinically acceptable scan times (5:30 min), with AMOCO providing image quality comparable to navigator-gated acquisitions. For flow measurements, AMOCO results showed insignificant net flow differences and improved consistency with 2D flow (ICC: 0.96 for AMOCO vs. 0.90 for uncorrected reconstructions). The proposed method is feasible for 4D flow motion correction across scanning platforms.
Conclusion:
AMOCO effectively suppresses motion artifacts, improves small-vessel visualization and accuracy of flow measurements while maintaining clinically acceptable scan times. Its potential for applications across different field strengths was preliminarily validated.
