A fast image registration algorithm for Digital Subtraction Angiography

Abstract

Digital Subtraction Angiography (DSA) is a powerful technique for the visualization of blood vessels. The major problem encountered in DSA images is the presence of motion artifacts which arise from the misalignment of successive images in the sequence. In this paper, a fast registration algorithm for coronary angiographic images is proposed. It involves a feature-based selection of control points for which the displacement is computed by means of template matching, and the complete displacement vector field is constructed by means of interpolation. To reduce the computational cost of the template matching procedure, an iterative hill-climbing approach is used for optimization and for handling the local misalignments, the image warping is performed based on multilevel B-spline interpolation.