Sudden Death Prediction Using Deep Learning
Implantable cardioverter-defibrillators (ICD) are the standard of care for primary prevention of sudden cardiac death (SCD) in high-risk patients. Currently, the decision for ICD implantation is guided by patient’s symptoms and heart function. However, only a minority of patients experience appropriate ICD firings. Cost, risk of complications, and inappropriate firings do not warrant indiscriminate implantation of ICDs. Our objective is to identify patient-specific markers of appropriate firing in ICD recipients using deep learning based on cardiac MRI acquired prior to ICD implantation.