Impact of acoustic window on accuracy of longitudinal global strain: a comparison study to cardiac magnetic resonance

Abstract

To evaluate the impact of acoustic window on the feasibility and accuracy of longitudinal global strain (global-ε) by speckle tracking for assessing left ventricular (LV) systolic function. The study included 70 patients (57 ± 17 years, 64% men), 28 selected patients with a suboptimal image quality (IQ) defined by three or more segments (4 ± 3 segments/patient) with wall motion score not analysable visually and 42 patients with an optimal two-dimensional (2D) echocardiography IQ. Left ventricular ejection fraction (LVEF) by Simpson's biplane method (2D-EF), global-ε by speckle tracking, and peak systolic mitral annulus velocity [systolic tissue Doppler imaging (S-TDI)] were compared with LVEF by cardiac magnetic resonance (EF-CMR; 45 ± 18%, range 9-76%). Speckle-tracking analysis was feasible in all segments with an optimal acoustic window and in 85% (103/121) of segments poorly visualized. Global-ε similarly correlated with LVEF by CMR in patients with and without optimal IQ (r = 0.81 vs. 0.82 for good vs. poor IQ). In contrast, 2D-EF (r = 0.76) and S-TDI (r = 0.64) less correlated with LVEF by CMR in patients with a suboptimal IQ. Importantly, IQ only impacted on 2D-EF inter-observer reproducibility (9 ± 5 vs. 24 ± 22% for good vs. poor IQ) but not on global-ε reproducibility (9 ± 1 vs. 8 ± 7% for good vs. poor IQ). In patients with a limited acoustic window, longitudinal strain by speckle tracking remains accurate and reproducible for assessing global and regional LV systolic function.