# Feasibility, reproducibility, and prognostic value of a fully automated measurement of left ventricular longitudinal strain in heart transplant recipients

## TL;DR

This study shows that automated software can quickly and reliably measure heart function in transplant patients and predict potential health issues.

## Contribution

The study demonstrates the feasibility and prognostic value of fully automated strain measurement in heart transplant recipients.

## Key findings

- Fully automated LVGLS measurement was feasible in 96.8% of patients and significantly faster than manual methods.
- Automated LVGLS could detect abnormal heart function and predict adverse events in transplant recipients.
- Semiautomated methods showed stronger correlation and smaller disagreement with manual measurements compared to fully automated methods.

## Abstract

Left ventricular longitudinal strain (LVGLS) is a robust parameter for predicting adverse events in patients who have undergone a heart transplant (HTx). However, measuring LVGLS is time-consuming and operator-dependent. Thus, we investigated whether automated strain software for LVGLS measurement has feasibility, reproducibility, and prognostic value in patients who underwent an HTx.

In total, 286 patients who had undergone heart transplants and comprehensive echocardiography were included. LVGLS was obtained from the same apical images by three different methods, namely, fully automated LVGLS measurement (Auto-Strain), semiautomated LVGLS measurement (automated with manual editing), and manual LVGLS measurement. Patients were followed up and the primary composite endpoint (defined as all-cause death and rejection) was recorded.

Fully automated measurements were feasible in 277 subjects (96.8%). Analysis times for the automated LVGLS (27.7 ± 2.8 s/patient) and the semiautomated LVGLS measurement methods (237.4 ± 41.0 s/patient) were shorter than for the manual LVGLS measurement method (440.4 ± 65 s/patient). The semiautomated LVGLS measurement method showed a stronger correlation with the manual LVGLS measurement method than the automated LVGLS measurement method (r = 0.854 vs. 0.654, P < 0.001), and there were smaller disagreements between the semiautomated LVGLS and manual LVGLS measurement methods [bias: 0.79, limits of agreement (LOA): 2.78] than between the automated LVGLS and manual LVGLS measurement methods (bias: 2.72, LOA: 3.98). During a median follow-up of 51 months (35.0–66.5 months), 35 patients experienced endpoint events. The automated LVGLS measurement method can detect abnormal systolic function and predict adverse events in patients who have undergone an HTx, while the detection and prediction ability of semiautomated the LVGLS measurement method was greater.

Fully automated LVGLS measurement enables rapid and reproducible assessment of graft function in patients who have undergone an HTx. Furthermore, the automated LVGLS measurement method detected abnormal systolic function and predicted adverse events, while the semiautomated LVGLS measurement method performed better in these aspects.