0333: Variability of right ventricular strain derived from speckle-tracking analysis using two different software solutions

Abstract

Speckle tracking imaging is a recent technique that can be achieved using either vendor dependent or vendor-independent software. Right ventricular (RV) strain is increasingly used as a prognostic tool in both left and right ventricular diseases. Only little is known regarding the variability of vendor-dependent and - independent speckle-tracking imaging software in the assessment of RV free wall longitudinal strain (RLS). The aim of our study was to compare a vendor-dependent (Qlab 9.0, Philips Medical System, Andover, MA, USA) and - independent (Cardiac Performance Analysis, Tomtec Imaging Systems, Germany) software for RLS analysis. We prospectively enrolled 90 consecutive patients with pulmonary hypertension (mean age 55,8±19years) and 26 control patients (mean age 33,9±13years) who underwent a comprehensive echocardiogram including a RV focused 4-chamber view optimised for speckle-tracking analysis. DICOM data sets were stored and analysed by 2 different cardiologists using Qlab and TomTec, blindly to the context and each other. In the whole population, mean RLS was -17,3%±9 and -8,6% ±7,2 respectively using Qlab and Tomtec. Qlab and Tomtec intra-observer coefficients of variation (CV) were -13,19% and -9,56% and interobserver CVs were -22% and -15% respectively. The concordance correlation coefficient was 0,55, indicating poor agreement between the two methods. In the control population, Qlab CV was - 3,63%, whereas CV was -17,8% in RV disease patients. Despite an acceptable level of variability for both techniques, Tomtec appears less variable. Variability of Qlab is excellent in control patients but is highly influenced by RV morphology. The agreement between the two software products is low and should lead in clinical practice to the follow-up of patients with the same software and advocates for the development of dedicated RV speckle-tracking software products.