A Results Recapitulation of Image Registration Techniques in Digital Subtraction Angiography

Digital Subtraction Angiography (DSA) is a powerful technique for the visualization of blood vessels in a sequence of X-ray images. In this application, it is necessary to register multiple images of the same sequence in order to reduce patient motion artifacts. The registered images are then subtracted in order to evaluate the progress of a disease and/or a treatment procedure. In this paper, a results recapitulation of some recent automatic registration techniques for digital X-ray angiography images is presented. A first method is based on a 3-D space-time motion evaluation and on an edge-detection approach leading to the selection of a set of control points from which pixels having their values changed due to motion are separated from those due to contrast flow, by means of detecting the peak iodine opacification. A second technique is based on local similarity detection by means of template matching according to the combined invariants-based similarity measure. The main characteristic of the third approach is that it combines the two previous ones to establish correspondences between the potentially matched movement points detected from the two images. The results clearly indicate that the third registration algorithm is much better able to recover patient motion than the other ones and would therefore be the method of choice for automatic registration of X-ray angiography images.