Comparison of the Prognostic Performance of Various Machine Learning Models in Patients with Acute Myocardial Infarction: Results from the COREA-AMI Registry
Analysis of various machine learning techniques shows random forest excelled in predicting major adverse cardiac events in AMI patients, highlighting clinical implications.
Key Points
The random forest model achieved the highest predictive performance for major adverse cardiac events in patients with acute myocardial infarction.
At 5 years, the random forest model had an area under the curve of 0.822 and an accuracy of 0.804, outperforming other models tested.
Analysis of predictors for adverse outcomes identified key factors including age, renal function, and adherence to optimal medical therapy.
This study emphasizes the ongoing role of guideline-directed medical therapy in improving prognosis for patients with acute myocardial infarction.