Comparison of cardiac strain parameters in healthy volunteers using multi-vendor speckle-tracking echocardiography

Aim        To determine and compare myocardial strain parameters for the left ventricle (LV), left atrium (LA), and right ventricle (RV) in healthy individuals using speckle-tracking echocardiography (STE) across equipment from two different manufacturers.Material and methods    The study included 30 subjects with no known cardiovascular disease at the time of enrollment. Transthoracic echocardiography (TTE) was performed on all participants on the same day within a single hour by the same experienced clinician, using two ultrasound systems from different manufacturers sequentially. To evaluate chamber strain, dedicated cine loops were recorded for the LV, LA, and RV. The following parameters were analyzed: LV global longitudinal systolic strain and segmental longitudinal strain (17-segment model); LV radial and circumferential strain; RV free wall global and segmental longitudinal strain; and LA strain during the reservoir, conduit, and contractile phases.Results. The longitudinal strain values for the LV, RV, and LA obtained via STE in healthy individuals using software from two different manufacturers fell within the established normal ranges. A comparison of the strain metrics between the two software packages revealed significant differences in regional longitudinal systolic strain in 6 out of 17 LV segments, 1 out of 3 RV free wall segments, as well as in LV radial and circular strain, and LA global longitudinal strain during the contractile phase. Bland-Altman analysis demonstrated a mean bias exceeding 1 and wide confidence intervals for the metrics across the two software platforms.Conclusion         The longitudinal strain values for the LV, RV, and LA obtained via STE in healthy individuals fell within the established normal reference ranges, regardless of the software vendor. A comparison of strain parameters calculated using software from different manufacturers revealed significant differences in several metrics. Consequently, strain values obtained from different software manufacturers are not directly comparable.