Diagnostic accuracy of 4D flow MRI comparing 2mm3 and 3mm3 spatial resolution. (8th February 2021)

• Record Type: Journal Article
• Authors: Cave, DGW; Shelley, D; Michael, H; Garg, P; Greenwood, JP; Plein, S; Olaru, MA; Van Der Geest, RJ; Bissell, MM
• Is Part Of: European heart journal. Volume 22(2021)Supplement 1
• Publication Date: 2021-02-08
• Subjects: Cardiovascular system -- Imaging -- Periodicals; Heart -- Imaging -- Periodicals; 616.10754
• DOI: 10.1093/ehjci/jeaa356.284
• Languages: English
• ISSNs: 2047-2404

Abstract

Abstract: Funding Acknowledgements: Type of funding sources: Public Institution(s). Main funding source(s): National Institute for Health Research, UK University of Leeds Background: Cardiac magnetic resonance (CMR) examinations requiring repeated breath-holds are challenging in younger patients. While 4-dimensional phase-contrast (4D flow) CMR does not require breath-holds, acquisition has been lengthy. Therefore to date spatial resolution has been influenced mainly by scan length. With accelerated sequences becoming available, higher spatial resolution is becoming clinically feasible. Purpose: We therefore evaluated the minimum spatial resolution in 4D flow CMR necessary for accurate clinical assessment. Methods: Ten healthy volunteers (mean age 24.8 years) underwent cardiac examinations on a 3T scanner using a 4D Flow prototype sequence at 2x2x2mm3 (4DFlow2) and 3x3x3mm3 (4DFlow3) spatial resolution. Net forward flow (FF) and peak velocity (PV) using valve tracking were calculated with commercially available software and kinetic energy (KE) in the left ventricle (LV) was analysed using a research tool. Bland-Altman analysis was used for statistical assessment and is reported as bias ± limits of agreement. Results: Aortic valve flow metrics were similar in 4DFlow2 (FF 94ml; PV 133cm/s) and 4DFlow3 (FF 95ml; PV 130cm/s), and both showed good agreement with 2D PC MRI (FF 93ml, Bland-Altman:1.6 ± 9.7 and 2.2 ± 13.5, respectively). Similar results were obtained for pulmonary valve flow (FF 138cm/s; Bland-Altman:4.7 ± 15.1 and 8.1 ± 18.2, respectively). Branch pulmonary artery (PA) FF showed good agreement with the main PA FF in 2D and 4DFlow2 (Bland-Altman:1.1 ± 15.9 and 1.1 ± 10.6, respectively), but not in 4DFlow3 (Bland-Altman:1.1 ± 32.5). Global LV KE measured by 4DFlow3 was on average 12% lower compared to 4DFlow2, whereas maximum systolic LV KE was similar in both acquisition methods. Conclusions: 3mm3 spatial resolution appears to be sufficient for clinical evaluation of aortic and pulmonary valves. Smaller vessels such as branch pulmonary arteries require higher resolution for accurate assessment. While no gold standard is available for kinetic energy assessment, our results suggest that some parameters LV energetic assessment is spatial resolution sensitive. Differences in SNR might also contribute to the differing results.