[Artificial intelligence for the diagnosis of acute coronary syndromes]

Artificial intelligence (AI) is assuming an increasingly prominent role in the diagnosis and management of acute myocardial infarction. Its main objective is to enable earlier and more accurate diagnosis, enhance the interpretation of the ECG, accelerate reperfusion times, and ultimately improve patient outcomes. The ECG represents an ideal substrate for the application of deep learning, owing to the vast availability of digital tracings, the association with confirmed diagnoses, and the inclusion of numerous clinical variables. Several systems even allow the automated analysis of photographs of paper-based ECGs, processed through deep learning algorithms. Current evidence indicates that: (i) in ST-elevation myocardial infarction, AI achieves sensitivity and specificity superior to those of experienced cardiologists, with an accuracy approaching clinical applicability; (ii) in non-ST-elevation myocardial infarction, clinical heterogeneity reduces diagnostic precision, yet AI still demonstrates significant discriminative power, serving as a valuable support tool for clinicians; (iii) emerging applications include the prediction of complete vessel occlusion and identification of the culprit coronary artery; and (iv) advanced algorithms may also estimate functional parameters such as ejection fraction and global longitudinal strain, thereby enriching prognostic stratification. In conclusion, AI applied to the ECG represents an innovative tool for the timely diagnosis of acute coronary syndromes. Its integration into clinical practice has the potential to support cardiologists both in confirming uncertain diagnoses and in rapidly selecting patients who should undergo revascularization.