Abstract
We propose a new nonlinear image registration model which is based on nonlinear elastic regularization and unbiased registration. The nonlinear elastic and the unbiased regularization terms are simplified using the change of variables by introducing an unknown that approximates the Jacobian matrix of the displacement field. This reduces the minimization to involve linear differential equations. In contrast to recently proposed unbiased fluid registration method, the new model is written in a unified variational form and is minimized using gradient descent. As a result, the new unbiased nonlinear elasticity model is computationally more efficient and easier to implement than the unbiased fluid registration. The unbiased large-deformation nonlinear elasticity method was tested using volumetric serial magnetic resonance images and shown to have some advantages for medical imaging applications.
Keywords
Source Code and Data
No source code files available for this publication.
Reviews
Grand roman Joldes
Monday 30 June 2008
In this paper the authors propose a registration method that combines a registration metric, a nonlinear elastic regularization and an unbiased registration constraint. The registration metric can be defined in terms of intensity matching or as mutual information. An additional unknown that approximates the Jacobian matrix is introduced in order to simplify the minimization process.
Although the authors say they are interested in parameter free registration methods, the proposed method has quite a few parameters (λ, β, ν, μ), whose selection criteria is not described in the paper.
One important part missing from this paper is the validation of the proposed algorithm (accuracy, convergence) using quantitative results. The qualitative results presented do not give any indication about the performance of the algorithm.
Paragraph 2.1: some references describing the metrics would be useful, especially for the mutual information metric derivation and computation of the intensity distributions.
Paragraph 2.2: the regularization term might not correctly describe the elastic behavior of the organ. For example, an organ could have sections with different material properties (both for the stored energy function and for the elastic material constants). In such a case the regularization term might introduce more errors instead of making the solution more accurate.