- An AI algorithm developed to predict the survival and neurologic prognosis of cardiac arrest patients
- Possible to predict the prognosis of emergency cardiac arrest patients more accurately
A research paper of the AI Big Data Research Team consisting of Jeon Gi Hyeon (director of the Department of Cardiology) as the head of the Clinical Research Department of Mediplex Sejong Hospital and AI Big Data Center's head Kwon Jun Myeong (director of the Department of Emergency Medicine) was published in the April issue of Resuscitation
(IF: 5.863), one of the world's most prestigious journal.
This paper describes the development of a deep learning based AI algorithm that predicts the revival potential and neurological outcome of outpatient cardiac arrest cases. Based on the research on registration of outpatient cardiac arrest in Korea, the research team analyzed the entire data of outpatient cardiac arrest cases from 2012 to 2015 to develop the algorithm.
In detail, using the data until the point of emergency room arrival, such as age, gender, when and where the cardiac arrest occurred, whether there were any witnesses and cardiopulmonary resuscitation was performed by a passer-by, initial ECG results of the emergency medical service, and heartbeat resumption during transport, the prediction model was developed, and accuracy verification showed high accuracy of 95%.
"The research on AI focuses on developing tools that can help medical staff in the field, rather than investigating the nature of diseases," said Jeon Gi Hyeon. He added, "We believe that the AI algorithm developed this time will make it possible to predict the occurrence of brain injury and the survival of the patient as soon as the patient comes in receiving a CPR by 119 staff, and ultimately help to predict the patient's prognosis and determine the treatment plan."
Mediplex Sejong Hospital will continue to lead the research and development of AI technologies and platforms by conducting follow-up research to be used for patient safety as well as patient treatment in an actual situation.