Velocity vector imaging in the measurement of left ventricular twist mechanics: head-to-head one way comparison between speckle tracking echocardiography and velocity vector imaging.

BACKGROUND: Velocity vector imaging (VVI) is based on myocardial feature tracking, which incorporates speckle and endocardial border tracking and allows myocardial strains, strain rates, and velocities to be quantified. However, the accuracies of VVI-derived left ventricular rotation-associated values have not been validated, and the aim of this study was thus to perform a head-to-head comparison of twist-related values determined by VVI and speckle-tracking echocardiography (STE).

METHODS: Thirty-five patients with a wide spectrum of cardiac pathologies and 19 healthy subjects (a total of 54 subjects) were enrolled. Apical and basal short-axis images were obtained using GE Vivid 7 scanners at frame rates of 80 to 100 frames/s. Left ventricular rotation-related parameters were first obtained using EchoPAC version 7.0.1. For comparison purposes, the same basal and apical short-axis images were converted into uncompressed Digital Imaging and Communications in Medicine format and subsequently analyzed using Syngo version 3.0, which uses a VVI tracking algorithm.

RESULTS: Basal and apical peak rotation and peak twist determined using STE and VVI were well correlated (r = 0.80, P < .001, r = 0.87, P < .001, and r = 0.90, P < .001, respectively). With regard to rotation velocities, peak rotation and derotation velocities were moderately correlated at the basal (r = 0.70, P < .001, and r = 0.72, P < .001, respectively) and apical (r = 0.66, P < .001, and r = 0.51, P < .001, respectively) levels. Furthermore, twist and untwist velocities were moderately correlated between the two methods. However, the timings of peak rotation and derotation velocities and twist and untwist velocities were only weakly correlated at the basal and apical levels.

CONCLUSIONS: VVI is a feasible modality for assessing twist-related parameters. Although VVI agreed well with STE for most of the rotation-related parameters, poor concordance was found between the two methods for a few parameters. It is hypothesized that this discrepancy originates from the different tracking algorithms used. The findings indicate that caution should be exercised when interpreting or comparing twist-related parameters obtained using different echocardiographic devices with proprietary image analyzers.