Detection of Coronary Plaque and Sensing Risks Factors of Heart at early stages using various Image Processing and Segmentation Techniques

Abstract

Cardiac arrest known as myocardial infarction is one of the common disease occurring in the human heart, usually causing death. Currently some of the clinical modalities are insufficient to detect at early stages. This paper shows the early detection of plaque non-invasively through CCTA images. In this paper, noise removal is being done by enhancement technique followed by global thresholding with segmentation using K-means Clustering. Further ROI extraction by stenosis using canny gradient operator with non maximum suppression with threshold values 75 to 90 and hysteresis is done. The proposed mathematical model explains the rate of change in blood flow using fluid dynamic concept by Hagen Poiseuille law and vascular wall shear stress method for quantification of healthy and diseased coronary arteries. After that we have classified the levels of heart attack as initial, mild and severe using ANFIS(Adaptive Neuro Fuzzy Inference System) tool with membership function. Computer simulation results assist in predicting the risk factors of heart attack at an early stage