Analysis and Classification of Patients with Acute Myocardial Infarction by Using Nonlinear Mixed-Effects Modeling

Abstract

In this preliminary study, the nonlinear mixed-effects modeling-based methodology has been chosen to investigate and evaluate the possible implication of some clinical cofactors on the release of biomarker cardiac troponin T (cTnT) in patients with acute myocardial infarction (AMI) and ST-segment elevation (STEMI). The aim of the study consists of the identification of subclasses of STEMI patients with different characteristics and, potentially, different clinical or pharmacological needs. An ad-hoc mathematical model, describing the biomarker release process subsequent to AMI, has been devised and exploited to estimate typical parameter values, and to evaluate the impact of covariates on the cTnT release curve. Among all the available co-factors, the mixed-effect analysis has found dyslipidemia to be a statistically significant one. More specifically, it has highlighted a relevant effect on the model parameters related to cTnT clearance. By increasing the number of co-factors, and enlarging the patients dataset, this approach may be useful in the automatic categorization and to unravel potentially unknown interactions between cofactors in AMI patients.