Phenomapping Patients After Percutaneous Coronary Intervention Using Novel Machine Learning Techniques: Insights Into the Melbourne Interventional Group Registry

Abstract

Background: Patients undergoing percutaneous coronary intervention (PCI) form a heterogeneous group with a multitude of independent and codependent risk factors. Readmission is a key clinical metric; however, identifying these patients on an individual level has proven difficult. This study aimed to use machine learning (ML) techniques to phenotype postPCI patients, and then developed an interactive dashboard from which to identify individual risks of readmission and generate individualised survival curves.