(C-98) Automated Quantification of Pulmonary Artery Pulse Wave Velocity in Tetralogy of Fallot Patients Using Two-Dimensional Phase-Contrast Magnetic Resonance Imaging

Introduction:: Pulse Wave Velocity (PWV), or the speed of wave propagation in a vessel, can provide useful information with regards to vessel stiffness and arterial tone.[1] To explore a deep, non-invasive alternative for measuring pulmonary arterial PWV, this study utilized 2D Cardiac Phase-Contrast Magnetic Resonance Imaging (PC-MRI) in a population with known pulmonary arterial pathology, repaired Tetralogy of Fallot (TOF). We compared automated methods for calculating pulmonary artery (PA) PWV using existing clinical PC-MRI data from repaired TOF patients. The goal was to identify whether clinical phase contrast MRI images had adequate spatio-temporal resolution to measure accurately pulmonary arterial PWV values, typically ~2 m/s in healthy adults and ~4 m/s in pulmonary hypertensive patients.[2] We hypothesized that limited spatio-temporal resolution in 2D-PC-MRI may reduce the accuracy of velocity extraction and transit time definition, leading to lower precision in PWV calculations, especially for short arterial segments like pulmonary arteries.


Materials and Methods:: This study analyzed existing cardiac 2D-PC-MRI data from 4 patients with repaired Tetralogy of Fallot acquired on a clinical 1.5 T scanner with 20 phases and a V_enc of 150 m/s. PA-PWV was extracted from velocity curves obtained in two orientations: the right ventricular outflow tract (RVOT) and the in-plane pulmonary arterial bifurcation views (SEAGULL). Regions of interest (ROIs) were selected along the PA to obtain velocity curves (Fig.1). Two automated time-to-foot (TTF) detection algorithms were used to determine the foot of each velocity curve. The first method utilized the 20 and 80 percentile points to interpolate the foot location based on the slope of the velocity curve (Fig.1). The second method used a maximum-derivative approach to locate the discrete point prior to the largest velocity increase and used it to determine the foot. PWV was calculated by plotting the TTF and the distance between each point along the PA, where the slope of the linear regression was the PWV.


Results, Conclusions, and Discussions:: The PWV mean for the RVOT orientation was measured using the 80-20 and maximum- derivative methods, resulting in 0.28 ± 0.18 m/s and 0.25 ± 0.05 m/s respectively. The pulmonary artery bifurcation view, when analyzed using the 80-20 method, had a mean PWV of 0.46 ± 0.14 m/s, whereas it was 0.45 ± 0.08 m/s with the max derivative technique. The resolution of our method to detect PWV on arterial lengths of 49.4 ± 9.2 mm was determined as 1.15 m/s, suggesting that velocities greater than this limit cannot be measured.
This work shows that PWV can be measured in the PA using standard 2D-PC-MRI. Although the lower observed PWV values in our study compared to the literature may be physiologically normal in TOF patients, further investigations with larger sample sizes, higher spatiotemporal resolution and control subjects are necessary for confirmation. Our data shows that higher temporal resolution or longer distances between interrogations is needed to measure velocities >1.15 m/s, common in pulmonary hypertensive patients. Overall, this study provides a promising proof-of-concept for improving non-invasive methods of acquiring pulmonary artery pressures with a PWV analysis using an automated approach.


Acknowledgements (Optional): :

References (Optional): : [1] [1] P. Boutouyrie, M. Briet, C. Collin, S. Vermeersch, and B. Pannier, “Assessment of pulse wave velocity,” Artery Res, vol. 3, no. 1, pp. 3–8, Dec. 2008, doi: 10.1016/J.ARTRES.2008.11.002.
[2] J. Sanz, S. Prat-Gonzalez, F. Macaluso, V. Fuster, and M. Garcia, “155 Quantification of pulse wave velocity in the pulmonary artery in patients with pulmonary hypertension,” Journal of Cardiovascular Magnetic Resonance, vol. 10, no. S1, Oct. 2008, doi: 10.1186/1532-429X-10-S1-A56/METRICS.