Sudden cardiac death (SCD) poses a significant health care challenge with high annual incidence and low survival rates. Implantable cardioverter defibrillators (ICDs) prevent SCD in patients with poor heart function. However, the critical survival benefit afforded by the devices is accompanied by short and long-term complications and a high economic burden. Moreover, in using current practice guidelines of reduced heart function, specifically left ventricular ejection fraction (LVEF)≤35%, as the main determining factor for patient selection, only a minority of patients actually benefit from ICD therapy (\<25% in 5 years). There is an essential need for more robust diagnostic approaches to SCD risk stratification. This project examines the hypothesis that structural abnormalities of the heart itself, above and beyond global LV dysfunction, are important predictors of SCD risk since they indicate the presence of the abnormal tissue substrate required for the abnormal electrical circuits and heart rhythms that actually lead to SCD. Information about the heart's structure will be obtained from cardiac magnetic resonance imaging and used in combination with a number of other clinical risk factors to see if certain characteristics can better predict patients at risk for SCD.