D. Zikic, A. Kamen, N. Navab Natural Gradients for Deformable Registration IEEE Conference on Computer Vision and Pattern Recognition (CVPR) in San Francisco, CA, USA, June 13-18 2010. (bib)
	We apply the concept of natural gradients to deformable registration. The motivation stems from the lack of physical interpretation for gradients of image-based difference measures. The main idea is to endow the parameter space of displacements with a distance metric which reflects the variation of the difference measure between two displacments. This is in contrast to standard approaches which assume the Euclidean frame. The modification of the distance metric is realized by treating the displacement fields as a Riemannian manifold. In our case, the manifold is induced by the Riemannian metric tensor based on the approximation of the Fisher Information matrix, which takes into accout the information about the chosen difference measure and the input images. Thus, the resulting natural gradient defined on this manifold inherently takes into account this information. The practical advantages of the proposed approach are the improvement of registration error and faster convergence for low-gradient regions. The proposed scheme is generally applicable to arbitrary difference measures and can be directly integrated into standard variational deformable registration methods with practically no computational overhead.
	This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each authors copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.