@CONFERENCE\{IMM2007-05045,
    author       = "M. Vester-Christensen and S. G. Erbou and S. Darkner and R. Larsen",
    title        = "Accelerated {3D} image registration",
    year         = "2007",
    month        = "feb",
    keywords     = "Image Registration, Inverse Compositional optimization",
    booktitle    = "International Symposium on Medical Imaging 2007, San Diego, \{CA\}",
    volume       = "",
    series       = "",
    editor       = "",
    publisher    = "",
    organization = "",
    address      = "",
    url          = "http://www2.compute.dtu.dk/pubdb/pubs/5045-full.html",
    abstract     = "Image registration is an important task in most medical imaging applications.  Numerous algorithms have been proposed and some are widely used. However, due to the vast amount of data collected, for instance by a computed tomography (CT) scanner, most registration algorithms are very slow and memory consuming. This is a huge problem especially in atlas building, where potentially hundreds of registrations are performed. This paper describes an approach for accelerated image registration. A grid-based warp function proposed, parameterized by the displacement of the grid-nodes, is used. Using a coarse-to-fine approach the composition of small diffeomorphic warps results in a final diffeomorphic warp.  Normally the registration is done using a standard gradient-based optimizer, but to obtain a fast algorithm the optimization is formulated in an inverse compositional framework. By switching the roles of the target and the source volume, the Jacobian and the Hessian can be pre-calculated resulting in a very efficient optimization algorithm.  By exploiting the local nature of the grid-based warp, the storage requirements of the Jacobian and the Hessian can be minimized. Furthermore, it is shown that additional constraints on the registration, such as the location of markers, are easily embedded in the optimization. The method is applied on volumes built from {CT-}scans of pig-carcasses, and results show a three-fold increase in speed using the inverse compositional approach versus the traditional gradient-based method."
}