An Efficient Technique for Detection and Enhancement of Blood Vessels from Cardiac MRI

Abstract

Cardiovascular diseases, mainly caused by stenosis

or aneurysms of arteries, are the leading reasons of death

worldwide. Angiography, which is not only invasive but also

risky, painful and costly, is the common method for the

recognition of such diseases. However presently non-invasive

and radiation free cardiac MRI is being used to evaluate the

function and anatomy of the heart chambers, heart valves and

blood flow through major vessels. The morphology, formation

and plaque buildup conditions of blood vessels indicates

diseases such as heart attack, stroke, angina, arrhythmias,

heart failure etc. Thus, the vessels needs to be identified

properly and clearly. In this paper, cardiac MRI is used for

accurate extraction and enhancement of blood vessel,

emphasizing upon the minor and somewhat discontinuous

vessels. MRI technique in this field was less explored and is

very challenging because of poor contrast difference between

the vessels and organs and mentionable amount of noise. The

detection process uses both morphological operators and image

gradient characteristics. The result obtained by using

morphological operators has better effectiveness in terms of

noise reduction and accuracy.