Validation of time-resolved, automated peak transvalvular velocity tracking through the mitral valve using four-dimensional flow cardiovascular magnetic resonance

Paul Njoku, Ciaran Grafton-Clarke, Hosamadin Assadi, Rebecca C. Gosling, Gareth Archer, Marcus Flather, Vassilios Vassiliou, Pankaj Garg

Research output: Contribution to journalAbstractpeer-review


Doppler echocardiography (TTE) remains the imaging modality of choice for the assessment of mitral inflow and left ventricular diastolic function, despite its limitations. Four-dimensional flow cardiovascular magnetic resonance (4D flow CMR) offers time-resolved cross-sectional velocity data, which can be used to investigate transvalvular peak velocity through the mitral valve. This would not suffer from the in-plane motion and angle-dependence of pulse-wave echocardiography.

Objective We aim to validate a novel time-resolved, automated dynamic 4D flow CMR peak velocity tracking method for measuring the peak velocity of mitral inflow against TTE.

Method Patients recruited to EurValve programme (n=22) underwent TTE and 4D flow CMR. Peak E-wave and A-wave velocities were recorded. This work was done in collaboration with the industry leader in 4D flow CMR (PIE Medical Imaging). Transvalvular flow segmentation was done using established valve tracking methods and the generated 3D streamlines were investigated for seeking the peak velocity inside the left ventricular cavity during diastole. Reproducibility analyses were carried out in 10 cases.

Results The peak E-wave mitral inflow velocity was comparable between the novel 4D flow method and TTE (1.09 ± 0.29 m/s and 1.10 ± 0.37 m/s respectively; p=0.60). The mean A-wave peak velocity was also comparable across both methods (0.94 ± 0.40 m/s and 0.86 ± 0.29 m/s respectively; p=0.38). The automated 4D flow method also showed good correlation with TTE for both E-wave (r=0.54; p=0.01) and A-wave (r=0.55; p=0.03) with minimal and non-significant bias between the two modalities (bias=0.01 m/s; p=0.91 and −0.08 m/s; p=0.91). This novel automated method demonstrated excellent reproducibility (Coefficient of variability 2.67% for peak E-wave mitral inflow velocity; Coefficient of variability 1.93% for peak A-wave mitral inflow velocity).
Conclusion We present a novel automated time-resolved transvalvular peak velocity assessment solution that can be used clinically for mitral inflow assessment and would circumvent the limitations of pulse-wave doppler echocardiography. Future studies are warranted to explore the diagnostic and prognostic advantages of our novel automated technique for mitral inflow assessment.
Original languageEnglish
Article number16
Pages (from-to)A15-A17
Number of pages3
Issue number3
Publication statusPublished - 1 Nov 2021

Cite this