Pilot tone navigation for respiratory and cardiac motion‐resolved free‐running 5D flow MRI. Issue 2 (5th October 2021)

• Record Type: Journal Article
• Authors: Falcão, Mariana B. L.; Di Sopra, Lorenzo; Ma, Liliana; Bacher, Mario; Yerly, Jérôme; Speier, Peter; Rutz, Tobias; Prša, Milan; Markl, Michael; Stuber, Matthias; Roy, Christopher W.
• Is Part Of: Magnetic resonance in medicine. Volume 87:Issue 2(2022)
• Page Start: 718
• Page End: 732
• Publication Date: 2021-10-05
• Subjects: 5D flow MRI -- cardiac motion -- free‐running -- pilot tone -- respiratory motion; Nuclear magnetic resonance -- Periodicals; Electron paramagnetic resonance -- Periodicals; 616.07548
• DOI: 10.1002/mrm.29023
• Languages: English
• ISSNs: 0740-3194

Abstract

Abstract : Purpose: In this work, we integrated the pilot tone (PT) navigation system into a reconstruction framework for respiratory and cardiac motion‐resolved 5D flow. We tested the hypotheses that PT would provide equivalent respiratory curves, cardiac triggers, and corresponding flow measurements to a previously established self‐gating (SG) technique while being independent from changes to the acquisition parameters. Methods: Fifteen volunteers and 9 patients were scanned with a free‐running 5D flow sequence, with PT integrated. Respiratory curves and cardiac triggers from PT and SG were compared across all subjects. Flow measurements from 5D flow reconstructions using both PT and SG were compared to each other and to a reference electrocardiogram‐gated and respiratory triggered 4D flow acquisition. Radial trajectories with variable readouts per interleave were also tested in 1 subject to compare cardiac trigger quality between PT and SG. Results: The correlation between PT and SG respiratory curves were 0.95 ± 0.06 for volunteers and 0.95 ± 0.04 for patients. Heartbeat duration measurements in volunteers and patients showed a bias to electrocardiogram measurements of, respectively, 0.16 ± 64.94 ms and 0.01 ± 39.29 ms for PT versus electrocardiogram and of 0.24 ± 63.68 ms and 0.09 ± 32.79 ms for SG versus electrocardiogram. No significant differences were reported for the flow measurements between 5D flow PT and from 5D flow SG. A decrease in the cardiac triggering quality of SG was observed for increasing readouts per interleave, whereas PT quality remained constant. Conclusion: PT has been successfully integrated in 5D flow MRI and has shown equivalent results to the previously described 5D flow SG technique, while being completely acquisition‐independent.