Simulation-free prediction of atrial fibrillation inducibility with the fibrotic kernel signature

Tomás Banduc, Luca Azzolin, Martin Manninger, Daniel Scherr, Gernot Plank, Simone Pezzuto, Francisco Sahli Costabal

Abstract: Highlights•We introduce a novel machine learning classifier to predict atrial fibrillation.•The classifier includes all the information about the anatomy and fibrosis.•We can predict atrial fibrillation inducibility without running simulations.•The method opens the door for fast personalization of therapies.