Based on raw images of a patient's heart, a new artificial intelligence algorithm developed at Johns Hopkins University can predict more accurately than doctors if and when a patient will die of cardiac arrest, revolutionizing clinical decision-making and improving survival rates for sudden and fatal arrhythmias.

The research team is the first to use a neural network to create a personalized survival assessment for each heart patient. These risk measures provide high accuracy for 10 years of sudden cardiac death and when it is most likely to occur. The algorithm's predictions were far more accurate than doctors' and were validated in tests at 60 medical centers in the United States.

Source from:

Popescu DM, Shade JK, Lai C, et al. Arrhythmic sudden death survival prediction using deep learning analysis of scarring in the heart. Nat Cardiovasc Res 2022; 1: 334–343. https://doi.org/10.1038/s44161-022-00041-9