AI Discovers New ECG Biomarker to Predict Sudden Cardiac Death

Researchers have uncovered a new ECG biomarker that predicts sudden cardiac death. The discovery was made using deep learning to analyse heart data. It marks a significant step forward in cardiovascular medicine. The team combined a predictive model with a generative one to isolate risk signals in ECG readings. This revealed a distinctive pattern in lead aVL: a smooth, slurred terminal R wave replacing the usual sharp S wave. The biomarker also involves axis deviation, such as left axis deviation and poor R-wave progression.

Unlike traditional markers like intrinsicoid deflection or fragmented QRS patterns, this new signal is spread across multiple ECG leads. This suggests a broader myocardial process at play. The findings highlight how AI can uncover hidden details in cardiac data, offering fresh insights into complex heart conditions. The biomarker could improve screening and risk assessment for sudden cardiac death. It may also enable earlier interventions for at-risk patients. The research demonstrates how AI can enhance scientific discovery in cardiology.