The application of ICA to the X-ray digital subtraction angiography

The traditional enhancement of X-ray digital subtraction angiography (DSA) is to subtract the mask image and living image so as to remove the back-ground such as ribs, spine, cathers, organs and etc, and obtain the enhanced vessel trees. However, the DSA have serious motion artifacts, poor local contrast and noises, when subtraction technique is used, some tiny vessels are broken, and even disappeared when visualized. To attack the problem, we use independent component analysis instead of subtraction technique. This technique is proved to be very efficient to enhance vessels. Experimental results of simulated data and several clinical data show that the proposed method is robust and can obtain good vessel trees.