Association Between Artificial Intelligence-Detected Features on the ECG and Presence of Microvascular Obstruction

The presence of microvascular obstruction (MVO) following reperfusion with percutaneous coronary intervention (PCI) in the setting of ST-segment elevation myocardial infarction (STEMI) portends a poor prognosis. MVO is associated with adverse left-ventricular (LV) remodeling (1), increased major adverse cardiac events and mortality (2). MVO may arise from both intravascular and extravascular mechanisms including distal embolization of thrombus and microvascular dysfunction as well as compression of the microcirculation from interstitial edema as a consequence of ischemia-reperfusion injury. The gold standard for detecting MVO is gadolinium-based cardiac MRI (CMR) which is expensive, time consuming and not universally available at STEMI centers. Identifying MVO is clinically important to help identify high-risk patients who may require intensive medical therapy. Detecting MVO without CMR can provide valuable insight into the outcomes of clinical trials by identifying those high-risk patients who may be more likely to benefit from novel therapeutics.