Rausch Karen, Shiino Kenji, Putrino Anthony, Lam Alfred King-Yin, Scalia Gregory M, Chan Jonathan
Department of Cardiology, The Prince Charles Hospital, Chermside, Brisbane, QLD, 4032, Australia.
School of Medicine, Griffith University, Gold Coast, Australia.
Int J Cardiovasc Imaging. 2019 Mar;35(3):419-426. doi: 10.1007/s10554-018-1453-7. Epub 2018 Sep 18.
Left atrial (LA) strain is an emerging technique with potential applications including arrhythmia prediction in atrial fibrillation and early identification of atrial dysfunction. The aim of this study was to evaluate reproducibility of LA strain and strain rate (SR) using multi-vendor analysis software between novice and expert. For LA strain to be a reliable tool, the technique must be reproducible by observers with variable experience. Use of multi-vendor analysis software allows serial strain assessment when echocardiographic images are acquired using different vendors. Fifty subjects underwent 2D-Speckle tracking echocardiographic (STE) derived LA strain and SR analysis measured from apical four and two-chamber views. Three strain parameters of LA function were assessed: reservoir (S-LAs, SR-LAs), contractile (S-LAa, SR-LAa) and conduit (S-LAs-S-LAa, SR-LAe). Strain analyses were performed by 2 independent, blinded novice and expert observers using multi-vendor analysis software. Intraobserver and interobserver analyses were performed using intra class correlation coefficients (ICC) and Bland-Altman analysis. LA strain and SR measured by novice observer demonstrated excellent intraobserver reproducibility (ICC for all strain and SR values > 0.88). There was good interobserver agreement of LA strain values between novice and expert (S-LAs:ICC 0.81, S-LAe:ICC 0.82, S-LAa:ICC 0.74). SR values also demonstrated good interobserver agreement (SR-LAs:ICC 0.83, SR-LAe:ICC 0.79, SR-LAa:ICC 0.86). Of all parameters, SR-LAa had the best interobserver and intraobserver agreement (ICC 0.86, 0.96). Global LA strain and SR values were highly reproducible by novice strain reader using multi-vendor analysis software. Interobserver reproducibility between novice and experts were good and acceptable within limits of agreement.
左心房（LA）应变是一项新兴技术，其潜在应用包括预测心房颤动中的心律失常以及早期识别心房功能障碍。本研究的目的是使用多厂商分析软件评估新手和专家之间LA应变及应变率（SR）的可重复性。要使LA应变成为可靠的工具，该技术必须能被经验各异的观察者重复使用。当使用不同厂商的设备获取超声心动图图像时，使用多厂商分析软件可进行连续应变评估。五十名受试者接受了二维斑点追踪超声心动图（STE）衍生的LA应变和SR分析，测量来自心尖四腔和两腔视图。评估了LA功能的三个应变参数：储存功能（S-LAs，SR-LAs）、收缩功能（S-LAa，SR-LAa）和管道功能（S-LAs-S-LAa，SR-LAe）。应变分析由两名独立、不知情的新手和专家观察者使用多厂商分析软件进行。使用组内相关系数（ICC）和布兰德-奥特曼分析进行观察者内和观察者间分析。新手观察者测量的LA应变和SR显示出极好的观察者内可重复性（所有应变和SR值的ICC>0.88）。新手和专家之间LA应变值的观察者间一致性良好（S-LAs：ICC 0.81，S-LAe：ICC 0.82，S-LAa：ICC 0.74）。SR值也显示出良好的观察者间一致性（SR-LAs：ICC 0.83，SR-LAe：ICC 0.79，SR-LAa：ICC 0.86）。在所有参数中，SR-LAa的观察者间和观察者内一致性最佳（ICC 0.86，0.96）。新手应变读取者使用多厂商分析软件时，整体LA应变和SR值具有高度可重复性。新手和专家之间的观察者间可重复性良好，在一致性范围内是可接受的。
