Application Progress of Machine Learning in Prognostic Prediction of Percutaneous Coronary Intervention: A Systematic Review

Abstract

Percutaneous coronary intervention (PCI) has become a cornerstone treatment for coronary artery disease; however, accurate prognostic prediction remains a significant clinical challenge. Machine learning technologies demonstrate tremendous potential in PCI prognostic prediction by analyzing vast amounts of clinical data and identifying complex nonlinear relationships among variables. This article systematically reviews recent advances in machine learning applications for PCI prognostic prediction, encompassing predictive targets including in-hospital mortality, major adverse cardiovascular events, bleeding complications, and long-term survival. Algorithms such as random forest, support vector machines, neural networks, and deep learning have demonstrated superior predictive performance compared to traditional risk scoring systems across multiple studies. Deep learning approaches exhibit particular advantages in processing multimodal data. Nevertheless, significant challenges remain regarding model interpretability, external validation, and clinical implementation. Future research should prioritize the development of explainable artificial intelligence systems, the conduct of multicenter validation studies, and the establishment of regulatory frameworks for clinical deployment.