Automated estimation of mitral annular plane systolic excursion by artificial intelligence from 3D ultrasound recordings

Anders Austlid Taskén, Erik Andreas Rye Berg, Bjørnar Leangen Grenne, Espen Holte, Håvard Dalen, Stian Stølen, Frank Lindseth, Svend Aakhus, Gabriel Kiss

Abstract: Highlights•Robust method for automatic and quantitative estimation of cardiac function.•Tested against reference measures provided by cardiologists for over 100 patients.•Automatic estimation of MAPSE in challenging 3D recordings, with large misalignments.•Landmark detection of the mitral annulus in 3D through the entire cardiac cycle.•Automatic alignment of ultrasound volumes to reduce misalignment and foreshortening.•Automatic detection of 2C-, 4C- and LAX views from cardiac ultrasound recordings.