Validation of Echodynamography in Comparison with Particle-image Velocimetry

Abstract

Echodynamography (EDG) is a computational method to estimate and visualize two-dimensional flow velocity vectors by applying dynamic flow theories to color Doppler echocardiography. The EDG method must be validated if applied to human cardiac flow function. However, a few studies of flow estimated have compared by EDG to the flow data were acquired by other methods. In this study, EDG was validated by comparing the analysis of estimating and visualizing flow velocity vectors obtained by original particle image velocimetry (PIV) based on a left ventricular (LV) phantom hydrogel (in vitro studies) and by EDG based on the virtual Doppler velocity. Velocity measured by PIV method and velocity estimated by EDG method in the perpendicular direction and the radial direction were compared. Regression analysis for the velocity estimated in the radial direction revealed an excellent correlation ($R^2=0.99$, slope = 0.96) and moderate correlation in the perpendicular direction ($R^2=0.44$, slope = 0.46). As revealed by the Bland-Altman plot, however, overestimations and higher relative error were observed in the perpendicular direction (0.51 ± 2.75 mm/s) and in the radial direction (-2.15 ± 21.13 mm/s). The percentage error of the norm-wise relative error of the velocity discrepancy is less than $10\%$, and velocity magnitude followed the same trends and are of comparable magnitude. These findings indicate that good estimates of velocity can be obtained by the EDG method. Therefore, the EDG method was appropriate for estimating and visualizing velocity vectors in clinical studies for higher measurement accuracy and reliability. The clinical in vivo application showed that the EDG method has the ability to visualize blood flow velocity vectors and differentiate the clinical information of vortex parameters both in normal and abnormal LV subjects. In conclusion, the EDG method has potentially greater clinical acceptance as a tool assessment of LV during the cardiac cycle.

Keywords: blood flow velocity; echodynamography; left ventricular; particle image velocimetry; phantom; vortex.