In myocardial infarction, the rapid identification of an acute coronary artery occlusion is critical to prevent heart muscle damage, yet current ECG criteria miss many cases. Here, the authors show that a deep learning model trained on over 500,000 ECGs can accurately identify and localize occlusion myocardial infarctions, outperforming standard criteria.
- Stefan Gustafsson
- Antônio H. Ribeiro
- Johan Sundström
