Speckle tracking analysis with the two main ultrasound software in pediatrics: Comparison between QLAB by Philips Healthcare and EchoPac by General Electric Healthcare

• Published in: Archives of Cardiovascular Diseases Supplements Vol. 10; no. 1; pp. 64 - 65
• Main Authors: Amedro, P., Bredy, C., Guillaumont, S., Lavastre, K., De La Villeon, G., Mura, T., Gamon, L., Cazorla, O., Lacampagne, A., Vincenti, M.
• Format: Journal Article
• Language: English
• Published: Elsevier Masson SAS 01.01.2018
• ISSN: 1878-6480
• DOI: 10.1016/j.acvdsp.2017.11.340

We aimed to compare, in a large cohort of normal children, ST analyses for right and left ventricles (RV, LV), between the QLAB 10.0 (Philips Healthcare) and the EchoPac 110.1.1 (General Electric Healthcare) ultrasound software. Normal children aged 0 to 18 years referred to the outpatient pediatric cardiology consultation were screened for this prospective cross-sectional study (ClinicalTrials.gov: NCT02056925). First echocardiography was performed after randomization on one of both ultrasound systems. Second echocardiography was performed with equal randomization upon 4 levels of variability: same ultrasound (intra-ultrasound variability), different ultrasound (inter-ultrasound variability), same sonographer (intra-observer variability) and different sonographer (inter-observer variability). Post-data ST analyses were performed twice by the same investigator (intra-analyser variability) and by a second investigator (inter-analyser variability). Hundred and fifty-six children were included. Mean age was 7.6±5 years [1 month–16.8 years]. There was no significant difference between both ultrasound systems in terms of conventional echocardiography variables. We observed a high proportion of missing ST data for the RV (19 to 78%). Inter-ultrasound system correlation was poor for both global longitudinal (Table 1) and circumferential LV strain, with an ICC of 0.34 (95% CI 0.06 to 0.57) and 0.12 (95% CI −0.18 to −0.40), respectively. We observed a very poor inter-observer reliability in both global LV longitudinal strain and global LV circumferential strain with the two software. Inter-analyzer variability was better especially for the global LV circumferential strain using Philips software (ICC 0.78 (95% CI 0.52 to 0.91). This study pointed out a serious limitation in the use of ST analysis in pediatrics with the two main ultrasound software, despite the EACVI/ASE Industry Task Force Standardization Initiatives for Deformation Imaging.