Reliability and reproducibility of trans-valvular flow measurement by 4D flow magnetic resonance imaging in acute myocardial infarct patients: two centre study

Abstract

Background: 4-dimensional (4D) flow MRI has advantage over 2-dimensional PC-MRI as it allows retrospective valve tracking, multi-planar dynamic phase contrast reconstruction to estimate stroke volumes (SV) reliably with good reproducibility in healthy subjects. However, this has not been demonstrated in patients with regional wall motion abnormality. We hypothesise that the SVassessed by 4D flow MRI through themitral valve (MV) and the aortic valve (AV) will be consistent. We also investigated mitral valve (MV) and aortic valve (AV)SVandalso the inter-observer reliability for intracardiacflowinacute myocardial infarction (AMI) patients. Methods: Fifteen patients underwent CMR at 1.5T (Ingenia CV, Philips Healthcare, Best, The Netherlands). CMR Protocol included: 2-chamber, 3-chamber, 4-chamber cines and 4DflowMRI with isotropic voxel size (3 × 3 × 3mm3), parallel imaging(SENSE2), velocity sensitivity Venc 150cm/s in all three directions and using echo-planar imaging (EPI) to factor of 5 for read-out acceleration. Free breathing was allowed and no respiratory motion correction was used. Retrospective gating was used and 30 cardiac phases were reconstructed. Images were analysed by two assessors (MH and PG) from two sites blinded to each other. Retrospective valve tracking with measurement planes positioned perpendicular to the inflow direction on 2-, 3- and 4-chamber cines was used to calculate SV. Background correction was used from velocity sampled in the myocardium. Results: MeanMVSV was 68±15 ml (PG) and 67±17 ml (MH) (p = 0.90). Mean AVSV was 70±20 ml (PG) and 73±14 ml (MH) (p = 0.96). Eight (53%) patients had mitral regurgitation, four (27%) had aortic regurgitation and four (27%) had tricuspid regurgitation. All regurgitation was graded as trivial/mild (MR fraction - 6.7±2%; AR fraction - 1.9±4 %). SV through MV and AV did not differ (PG: intraclass correlation coefficient (ICC) was 0.94, CI 0.81-0.97 p = 0.65; MH: ICC = 0.86, CI 0.46-0.96 p = 0.35). Coefficient of variation (CV) of inter-observer variability forMVSV was 6.3% and the concordance correlation coefficient was 0.93 CI (0.81-0.97) with accuracy of 0.99. CVof inter-observer variability for AV SV was 6.4% and the concordance correlation coefficient was 0.90 CI (0.72-0.96) with accuracy of 0.98. Conclusion: 4D flow MRI with retrospective valve tracking provides reliable assessment of transvalvular flow in AMI patients with high inter-observer agreement. (Figure Presented).