CORONARY ARTERY DETECTION IN X-RAY ANGIOGRAM WITH FUZZY C-MEANS CLUSTERING ALGORITHM

Abstract

This paper presents an algorithm for the segmentation of Blood Vessels from x-ray angiography of coronary artery images using Matched filters and Otsu thresholding . Identifying these arteries is a difficult process because of the vessel overlapping, and superimposition with various anatomical structures such as the ribs, the spine and the heart chambers. Firstly, the 3D Fourier and 2D Wavelet transforms are first used to eliminate the background and reduce noise in the images. Afterwards, a set of matched filters was applied to enhance the coronary arteries in the images. Finally with Otsu thresholding and 8-connected neighborhood pixels the segmented vessel was obtained.