Automated measurements based on artificial intelligence software in pediatric echocardiography: a feasibility study

Abstract Funding Acknowledgements Type of funding sources: None. Background Transthoracic echocardiography is the first-line non-invasive investigation for assessing pediatric patients' cardiac anatomy, physiology, and hemodynamics, based on its accessibility portability, but complete anatomic hemodynamic assessment time-consuming. The study aimed to determine accuracy reliability a new automated artificial intelligence software initially developed adults analyzing main echocardiographic parameters in patients. Materials Methods was performed at University Hospital Bordeaux between August September 2022 included 45 patients who underwent 2D transthoracic examination, mainly murmur (n=8), post ASD closure evaluation(N=5), chest pain, or arrhythmia as we have homogeneous group with normal near heart architecture. We compared AI automatic measurements manual by senior junior cardiologist. Pearson correlation Bland Altman analysis were used. As has been adults, decided subclassify our into two subgroups: first (patients aged 0–8 years) second over 8 years). decision made because anatomy differs from see if dedicated soft better reproducibility older than years. Results median age 8.9 Comparison cardiologists all subgroups significant (Pearson >0,7 most performed). For younger (<8 years old), found excellent correlations >0.8) chambers measured A4CH A2CH views (LV end-diastolic volume, LV end-systolic RV area, area) both cardiologist, poor parasternal long axis view (r=0.06 STJ measurement). achieved stronger evaluation old ( r>0.8 aortic annulus, STJ, E, A, S waves, volume A4CH, A4CH) cardiologist relation measurements. Conclusion In this preliminary study, that provided promising results measuring volumes volume) A2CH. requires improvements within long-axis section achieve accuracy. obtained significantly improved those under age.