Vendor and software based variation in left atrial strain measurements: implications for clinical practice.

Left atrial strain (LAS) is a valuable echocardiographic marker of left atrial function with growing clinical utility. However, variability in LAS measurements across software vendors remains a barrier to its routine clinical use. This study aimed to compare LAS measurements obtained using a dedicated left atrial-specific measurement tool (AFI-LA (General Electric (GE)) with those derived from established LV-based strain platforms: GE EchoPAC (mid-myocardial and endocardial tracking) and TomTec-Arena (feature tracking), used for measurement of LAS. LAS was measured in 112 subjects (72 individuals in sinus rhythm and 40 paroxysmal atrial fibrillation (PAF) patients in sinus rhythm at the time of transthoracic echocardiography). Reservoir (ƐR), contractile (ƐCT), and conduit (ƐCD) phasic strain were measured using AFI-LA and compared with GE-mid, GE-endo, and TomTec strain measurements. Agreement was assessed using Bland-Altman analysis and Pearson correlation. Inter- and intra-observer reproducibility was evaluated. AFI-LA measurements showed good correlation with both platforms (r ≥ 0.7, p < 0.001). AFI-LA consistently underestimated strain values across all phases compared to GE-mid, GE-endo, and TomTec, with the smallest bias observed against GE-mid. Proportional bias was present, particularly at higher strain values. Inter- and intra-observer reproducibility of AFI-LA measurements was high (r > 0.85). AFI-LA provides LAS measurements that are most comparable to GE mid-myocardial strain but demonstrates systematic underestimation compared to GE endocardial and TomTec-derived values. These differences underscore the need for vendor-specific calibration before AFI-LA can be used for serial assessments or applied using existing clinical threshold values. Larger validation studies are needed to support standardization and broader clinical adoption.