Feasibility, reproducibility, and clinical implications of the fully automated assessment for global longitudinal strain
Abstract Background Despite of evidence on its usefulness, measurement of global longitudinal strain (GLS) has not been widely accepted as a clinical routine because it requires proficiency and is time-consuming. Automated assessment of GLS may be a solution to these barriers. This study sought to investigate the feasibility, reproducibility, and predictive value of automated strain analysis compared with semi-automated and manual assessment for global longitudinal strain. Methods In this validation study, different methods for the assessment of GLS were applied to echocardiograms of 561 asymptomatic people (age 71±5 years) with heart failure (HF) risk factors, recruited from the community. All patients were followed up for new-onset of HF and cardiovascular death. Measurement of GLS was repeated using the same apical images on three different measurement packages as follows: (1) fully automated GLS (AutoStrain), (2) semi-automated GLS (automated, corrected by a trained investigator), and (3) manual GLS (standard manual assessment by a trained investigator). We defined abnormal GLS for discrimination of LV systolic dysfunction using the cut-off of GLS =18% (absolute value). Results AutoStrain measurements were feasible in 99.5% of patients. Calculation time for automated (0.5±0.1 min/patient) and semi-automated assessment (2.7±0.6 min/patient) were significantly shorter than that for manual assessment (4.5±1.6 min/patient) (both p<0.001), and the automated assessment showed excellent reproducibility. There was considerable discordance between automated and semi-automated/manual GLS (Figure 1), but normal systolic function was reliable identified. The prediction of cardiovascular events was reliable with automated, semi-automated and manual GLS (Figure 2). Conclusion A novel fully automated assessment for GLS is a feasible, rapid, reproducible and clinically applicable means of assessing LV function, and measurements in the normal range predict a favorable outcome. Figure 1 Funding Acknowledgement Type of funding source: None