Virtual Booth

Coronary artery disease (CAD) occurs when coronary arteries become stenotic or blocked, reducing blood flow to the heart and potentially leading to heart attacks. Diagnosis is commonly performed using coronary angiography (CAG) and analyzed with quantitative coronary analysis (QCA). However, QCA can be time-consuming. This study aims to develop an automated program to detect vessel boundaries in CAG images, calculate stenosis percentage, and simulate blood flow behavior in stenotic arteries to aid diagnosis and treatment. The methodology involves image processing and numerical simulations. Vessel boundaries are extracted using thresholding and flood fill, followed by K-means clustering to define stenotic regions. The stenosis percentage is calculated based on vessel width. Blood flow is simulated using the Navier-Stokes equations, with computations divided into three steps: (1) estimating momentum and viscosity using finite difference approximation, (2) calculating pressure via Poisson’s equation and the Gauss-Seidel method, and (3) determining blood velocity using the Euler method. Experimental results showed that blood velocity increased while pressure decreased in stenotic areas, aligning with fluid dynamics principles. The accuracy of stenosis calculations depended on CAG image quality—high-resolution images improved precision, whereas noisy images caused errors. The choice of time step (dt) affected computational stability, with optimal dt values enhancing accuracy. The combined analysis of stenosis percentage and blood flow behavior provides a more comprehensive CAD assessment, supporting improved treatment planning.