Sudden Unexpected Death in Epilepsy (SUDEP) is a poorly understood phenomenon. Patient cohorts to power statistical studies in SUDEP need to be drawn from multiple centers due to the low rate of reported SUDEP incidences. But the current practice of manual chart review of Epilepsy Monitoring Units (EMU) patient discharge summaries is time-consuming, tedious, and not scalable for large studies. To address this challenge in the multi-center NIH-funded Prevention and Risk Identification of SUDEP Mortality (PRISM) Project, we have developed the Epilepsy Data Extraction and Annotation (EpiDEA) system for effective processing of discharge summaries. EpiDEA uses a novel Epilepsy and Seizure Ontology (EpSO), which has been developed based on the International League Against Epilepsy (ILAE) classification system, as the core knowledge resource. By extending the cTAKES natural language processing tool developed at the Mayo Clinic, EpiDEA implements specialized functions to address the unique challenges of processing epilepsy and seizure-related clinical free text in discharge summaries. The EpiDEA system was evaluated on a corpus of 104 discharge summaries from the University Hospitals Case Medical Center EMU and achieved an overall precision of 93.59% and recall of 84.01% with an F-measure of 88.53%. The results were compared against a gold standard created by two epileptologists. We demonstrate the use of EpiDEA for cohort identification through use of an intuitive visual query interface that can be directly used by clinical researchers.