Different vendors show comparably high reproducibility of 3D strain measurements except for radial strain

Background: Recent advances in speckle-tracking echocardiography make it possible to quantify the complex left ventricular (LV) myocardial deformation in its three-dimensional (3D) perspective. Little is known so far about the reproducibility of 3D strain parameters and almost exclusively from a single vendor’s platform. Aim: To assess and compare the reproducibility of 3D strain parameters obtained with two commercially available systems: Vivid E9 with 4V probe and EchoPac BT11 software (GE Healthcare, Horten, NO) and Artida with PST-25SX probe and 3D WMT software (Toshiba Medical Systems, Tokyo, JP). Methods and results: Twenty patients referred for routine echocardiographic study (59±16 years, 12 men, LV ejection fraction range 30-70%) were imaged with both scanners during the same examination. 3D data sets were analyzed for inter-observer reliability by two separate observers. Intra-observer measures were performed one week apart. Peak global longitudinal, radial, circumferential and area strain were measured. Bland-Altman analysis and intra-class correlations were performed, and results for E9 are displayed in Table 1. Artida showedsimilar reproducibility for all strain components, except for a higher variability of radial strain in comparison to E9: bias (±LOA)=0.5%(±15.6%), ICC=0.56 for intraobserver reproducibility, and bias (±LOA)=4.8%(±22.2%), ICC=0.44 for interobserver reproducibility (p<0.001 for both, respectively).