Personalized Risk Prediction of Sudden Cardiac Death

WE will test the hypothesis that interpretable artificial intelligence models including clinical, bio-humoral, genetic, electrocardiography, echocardiography, cardiac magnetic resonance, and nuclear imaging data will predict the risk of SCD in patients with left ventricular dysfunction more accurately than NYHA class and LVEF, and this will translate in a more cost-effective use of ICDs compared to current clinical practice.