Automatic correction of angle-dependent and angle-independent geometric distortions in angiographic images

Abstract

To avoid inaccuracies when performing quantitative Coronary Angiography (QCA), it is very important to correct the angiographic images for geometric distortions. These geometric distortions appear as a result of the properties (or in other words, the imperfections) of the X-ray system of the angiographic device, and the presence of a magnetic field around the X-ray system. Distortions caused only by the properties of the X-ray system are identical for all images, independent of the projection angles corresponding to the images, and can be corrected relatively easily. On the other hand, distortions caused by the presence of a magnetic field around the X-ray system do depend on the projection angles. Correcting an image for these angle-dependent distortions requires the knowledge of the projection angles corresponding to the image, and the distortion properties corresponding to this particular set of projection angles. Here, the authors present a methodology to correct angiographic images automatically for angle dependent and angle-independent geometric distortions, based on the methodology proposed in A. Wahle et al. (Proc. 18/sup th/ Ann. Int. Conf. of IEEE Engng. in Med. and Biol. Soc., vol. 2, p. 654-5, 1996/97). The presented method does not require additional acts or markers during the patient examinations, nor additional input of data while correcting the images.