British researchers at the University of Leicester have created an AI tool, VA-ResNet-50, that can predict a person’s risk of lethal heart rhythm with 80% accuracy. Ventricular arrhythmia is a serious condition that can lead to sudden death if not treated immediately.
The tool was used to analyze Holter electrocardiograms of 270 adults between 2014 and 2022, with 159 experiencing lethal ventricular arrhythmias. VA-ResNet-50 correctly predicted which patients were at risk in 4 out of 5 cases.
Current clinical guidelines are not accurate enough in determining the risk of ventricular arrhythmia, leading to a significant number of deaths. Professor Andre Ng believes that using AI to analyze ECGs while in normal rhythm can help save lives by identifying those at highest risk.
“If the tool said a person was at risk, the risk of lethal event was three times higher than normal adults,” Professor Ng stated. Using AI in this way provides a new approach to assessing risk and recommending appropriate treatment, ultimately saving lives.
