flâneur

An ECG biomarker for sudden cardiac death discovered with deep learning | Nature

nature.com · 9,018 words · saved by 1 readers

A deep-learning model trained on electrocardiogram (ECG) waveforms identifies an easily visible biomarker that predicts sudden cardiac death more accurately than the current clinical state of the art.

Main Every year, cardiac arrhythmias cause hundreds of thousands of sudden deaths in the USA alone1,5. These deaths occur despite the availability—since 1980—of implanted cardioverter defibrillators, which can detect and terminate arrhythmias before they kill. Implanting a defibrillator has costs, so choosing the right patients requires risk prediction: the higher the likelihood of future arrhythmia, the more the benefits outweigh the costs6,7,8,9. The current state of the art for risk prediction is a biomarker measured through cardiac ultrasound: the heart’s left ventricular ejection…

saved by

related reading