Comparison between semi-automatic and manual left atrial strain analysis in patients undergoing CABG: feasibility, reproducibility, and clinical implications.

Abstract

Background: Left atrial (LA) strain, a powerful parameter for assessing LA function, is underused in routine clinical practice because it requires expertise and time. The novel semi-automated method for the measurement of LA strain may circumvent these limitations. Hence, we investigated the feasibility, reproducibility, and clinical implications of the semi-automated method.

Methods: We prospectively enrolled 190 patients referred for coronary artery bypass grafting (CABG). LA strain was measured using the fully semi-automated and manual methods from the same apical four-chamber view 3 days before CABG. Patients were followed up every 6 months for major adverse cardiovascular and cerebrovascular events (MACCE).

Results: Manual border corrections were required in 26% of patients because of poor tracking quality. Measurement time was significantly shorter with the semi-automated method. Excellent intra- and inter-observer measurement agreement with minimal bias were observed for LA strain measured using the semi-automated method, which were better than those for LA strain measured using the manual method. Moreover, measured LA strain using the semi-automated method had a greater correlation coefficient with the LA ejection fraction and prediction power for left ventricular diastolic dysfunction. Over a median follow-up of 22.4 months, MACCE occurred in 36 (19.0%) patients. LA strain measured using both methods showed comparable usefulness for predicting MACCE.

Conclusions: The novel semi-automated method rapidly measures LA strain with excellent reproducibility and has clinical application comparable to the manual method. This method may have the potential to be integrated into routine clinical practice and facilitate the implementation of LA strain.

Clinical trial number: Not applicable.