Are peak velocities determined by 2D phase-contrast MRI comparable to those assessed by real-time phase-contrast MRI and pulse wave echocardiography?

Introduction: The aim of this study was to compare peak velocities (Vmax) assessed by two-dimensional (2D) phase-contrast velocity mapping (PC-MRI), real-time PC-MRI and pulse wave echocardiography regarding the degree of discrepancy.

Methods: Quantitative through-plane PC-MRI flow/Vmax measurements were performed in the ascending aorta of 11 healthy subjects (mean age = 42.6±13.7 years; 6 male) with a 3 T scanner. Using (a) retrospectively gated standard 2D PC-MRI,(b) 2D PC-MRI (comparable to a) but with flow encoding in 3 spatial directions for 3D vector calculating; (c) real-time PC-MRI; (d) pulse wave echocardiography. Measurements were performed without repositioning of the subject. Normally distributed data were analyzed by paired Student-t-test regarding a p-value of 0.05 as statistically significant, Bland-Altman statistics and by calculating of regression coefficients.

Results: Average Vmax(MR_3D) exceeded Vmax(MR_ref) by 4,4% (=mean, limits-of-agreement: +14.1% to -5.3%, r∼0.96, p<0.05). Peak velocities were higher applying real-time PC-MRI compared to MR_ref (28.1%; +51.4% to +4.9%, r∼0.64, p<0.05), which can be explained by the non-averaging character of this technique representing a snap-shot of actual blood flow. Accordingly, Vmax(echo) was increased by 25.2% in relation to Vmax(MR_ref) (+69.0% to -18.5%, r<0.5, p<0.05). Comparing real-time PC-MRI and pulse wave echocardiography high agreement was observed (2.7%; +44.5% to -39.2%, r<0.5, p<0.05) but with substantial scatters and a low correlation coefficient.

Conclusions: Vmax assessed by standard quantitative two-dimensional PC-MRI were underestimated by 25-30% in comparison to real-time PC-MRI and pulse wave echocardiography.