Optical edge detection of chest MRA using combined contrast enhancement algorithms

Abstract

Most animals see little or no color at all. The same fact applies to the Magnetic Resonance technique: the signals are located mostly in shadowy zones interrupted by few light zones. Therefore, the MRAs (Magnetic Resonance Angiograms) are hard to interpret by the physician. MRA offers an image of how the blood spreads through the vessels and organs of the body. Both physician and patient can see where the pathway followed by the blood is blocked. As prevention is better than curing, we focus on finding an algorithm to improve the image contrast and outline the regions of interest. We aim thus to allow an early detection of the illness. For our study, we applied a combined method on biomedical images, to improve their optical contrast: an edge detection algorithm and a strong Matlab contrast-enhancement method named Contrast Limited Adaptive Histogram Equalization. Thus, we should allow the detection of the vascular system or the edges of the organs and improve the chances of an accurate diagnosis. The resulted contrast improvements are visible, unmasking medical features (hidden through the low contrast of the image).