Investigation of edge detection techniques based on coronary angiography images
Loading...
Date
2017-06
Authors
Basaier Jialade
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Often, low quality medical X-ray images are produced due to the least possible
amount of projection used to form the images. Minimum projection beams are used to
reduce damage to patients’ body. Angiography is an important class of medical imaging.
The image is obtained using X-ray techniques after a contrast agent is injected into the
patient’s blood vessels. Coronary angiography is a critical type of angiography images
and it could be used to diagnose serious heart and artery diseases. However, most
angiography images suffer from non-uniform illumination and noises, which makes it
difficult for doctors to make accurate diagnoses. Therefore, processing of angiograms is
necessary to make them more visible thus help doctors to produce more accurate
diagnoses in a shorter time. In this project, the angiogram is firstly enhanced using
histogram equalization. Then five edge detection techniques including Roberts cross,
Prewitt, Sobel, Canny, and a modified Sobel algorithm are implemented. The main aim
is to clearly delineate edges of the blood vessels, and the Sobel operator is found to be
the most suitable algorithm for analyzing coronary angiography images based on the
quantitative and qualitative evaluations.