EP02.11: Accuracy of an artificial intelligence software (HEARTASSIST^™) in automatically assessing the quality of second trimester cardiac planes

Congenital heart disease (CHD) are the most frequent congenital anomalies, but their diagnosis is frequently missed during routine prenatal ultrasound screening. There is evidence that the quality of cardiac planes obtained by the sonographer play a major role in the detection rate of CHD and may require the supervision of an expert not always available in all the settings. In this study we e tested an artificial intelligence software in the assessment of the quality of standard fetal cardiac planes in a prospective study and we compare the agreement with the evaluation of an expert. In a prospective observational study frames of the four chamber view left and right outflow tracts, and three-vessel trachea view were obtained from 120 consecutive singleton low-risk women undergoing second trimester ultrasound. For each frame, the quality assessment was performed by an expert sonographer and by an artificial intelligence software (Heartassist™). The Cohen's κ coefficient was used to evaluate the agreement rates between both techniques. The number and percentage of images considered adequate visually by the expert or with Heartassist™ were similar with a percentage > 87% for all the cardiac planes considered. The Cohen's κ coefficient values were for the four chamber view 0.827 (95% CI 0.662-0.992), 0.814 (95% CI 0.638-0.990) for left ventricle outflow tract, 0.838 (95% CI 0.683-0.992) and three vessel trachea view 0.866 (95% CI 0.717-0.999), indicating a good agreement between the two techniques. Heartassist™ allows to obtain the automatic evaluation of fetal cardiac planes, resulted reproducible, reached the same accuracy of expert visual assessment and has the potential to be applied in the evaluation of fetal heart during second trimester ultrasonographic screening of fetal anomalies.