Simulating Atherosclerotic Plaque Growth in Coronary Arteries Using an AI-Driven ODE Model for Personalized Cholesterol-Driven Risk Dynamics

Abstract This study introduces a simulation-based framework to model atherosclerotic plaque growth in coronary arteries, a critical factor in cardiovascular disease. The model uses a first-order ODE solved with MATLAB’s ode45 (a method for solving Equation ions) to show how plaque grows, which is affected by LDL cholesterol and blood pressure. For a fictitious group of 500 patients, a neural network uses characteristics such as age, LDL cholesterol, and blood pressure to forecast individual ODE parameters. Simulations over 15 years produce personalized plaque trajectories, enabling risk stratification into low, medium, and high-risk groups. MATLAB-created visuals, like time-series graphs, cholesterol-plaque scatter plots, and 3D sensitivity surfaces, provide a clear understanding of how diseases progress and how different factors affect them. The model has been checked against known solutions, showing it is very accurate and highlighting how AI and math can work together in health informatics to improve personalized heart disease risk assessment and precision medicine.