A numerical study of ultrafast vector flow imaging in the neonatal heart

Abstract

Doppler flow imaging for the visualisation of neonatal intraventricular blood flow currently still has inherent limitations: beam-to-flow angle dependency, aliasing and a too low frame rate. Ultrafast imaging and vector flow estimation may resolve these limitations, yet both still require thorough validation for the pediatric cardiac setting. Hence, a computational modelling approach in the neonatal left ventricle was employed to investigate (i) diverging wave emission to acquire images at very high frame rate and (ii) subsequent speckle tracking algorithms for vector flow estimation. Single non-tilted diverging waves with an opening angle of 60° were transmitted, at a pulse repetition frequency of 9 kHz. Speckle tracking on the acquired ultrasound images provided 2D intraventricular flow estimates at a frame rate of 180 Hz for both the apical four chamber and parasternal short axis view, and this over an entire cardiac cycle. Overall, the blood flow was reasonably accurately tracked throughout the cardiac cycle, yet several imaging artefacts were observed. Zones of low flow proved very difficult to track due to clutter filtering issues, while high spatial flow gradients caused strong underestimation of systolic outflow