Multiscore Integration of GRACE, ACEF, and TIMI via Machine Learning for Risk Prediction of 6-Month Mortality in Acute Myocardial Infarction
This study demonstrates that a machine learning model integrating variables from GRACE, ACEF, and TIMI scores, optimized via a dual-stage feature selection process, achieves excellent discrimination for predicting 6-month mortality in acute myocardial infarction patients undergoing PCI, ultimately identifying a parsimonious three-variable model (Killip class, LVEF, and creatinine) that balances simplicity with robust clinical performance.