Machine Learning For Ecg Diagnosis And Risk Stratification Of Occlusion Myocardial Infarction

This study describes the first multi-site, prospective, observational cohort study to evaluate the diagnostic accuracy of machine learning for the ECG diagnosis and risk stratification of OMI at first medical contact and in the absence of a STEMI pattern. The results demonstrate the superiority of machine learning in detecting subtle ischemic ECG changes indicative of OMI in the absence of a STEMI pattern, outperforming practicing clinicians and other widely used commercial ECG interpretation software. The study identified the most important ECG features driving the model’s classifications and identified plausible mechanistic links to myocardial injury. A new OMI risk score was derived, providing enhanced rule-in and rule-out accuracy when compared to the HEART score.