AI Captures ECG Patterns to Predict Future Sudden Cardiac Arrest

Now, two new studies by investigators at Cedars-Sinai (Los Angeles, CA, USA) support the use of AI in sudden cardiac arrest prediction. The first study involved training a deep learning algorithm to analyze electrocardiogram (ECG) patterns, which are recordings of heart electrical activity. The model examined ECGs from individuals who had suffered sudden cardiac arrest and those who had not, including 1,827 pre-cardiac arrest ECGs from 1,796 individuals who later experienced sudden cardiac arrest, and 1,342 ECGs from 1,325 people who did not. This Cedars-Sinai-developed AI model outperformed conventional methods, like the ECG risk score, in predicting out-of-hospital sudden cardiac arrest.

The second study focused on distinguishing between two causes of sudden cardiac arrest: pulseless electrical activity, where the heart’s electrical signals are too faint to produce a heartbeat, and ventricular fibrillation, an irregular heartbeat that can be treated with a defibrillator. After analyzing ECG patterns and patient characteristics, the researchers identified specific risk factors for each type. Patients with pulseless electrical activity sudden cardiac arrest were often older, overweight, anemic, or experienced shortness of breath. In contrast, those with ventricular fibrillation tended to be younger and had a history of coronary artery disease or chest pain as a warning sign.

“These studies exemplify the potential for AI to detect patterns in the body that the human eye and standard medical tests cannot,” said Paul Noble, MD, the Vera and Paul Guerin Family Distinguished Chair in Pulmonary Medicine and chair of the Department of Medicine at Cedars-Sinai. “We are getting closer to being able to use AI to prevent dangerous events such as sudden cardiac arrest.”

Related Links:
Cedars-Sinai