Diffusion-based Image Registration: How to Choose the Optimal Registration Parameters

J. Larrey-Ruiz, J. Morales-Sánchez, and R. Verdú-Monedero (Spain)


Diffusion Registration, NonRigid Matching, Optimal Pa rameters.


The task of image registration is to find an optimal geome tric transformation between corresponding image pairs. In practice, choosing the adequate registration and simulation parameters in non-parametric image registration methods, where the transformation is not restricted to a parameteriza ble set of mapping functions, is an open question. There is no agreement about which are the optimal values (if any) for these parameters, since they can also depend on the par ticular set of images to be registered. As a result, in the lit erature the parameters involved in the registration process are arbitrarily fixed by the authors. The present paper is intended to address this issue. A two-step method is pro posed to obtain the optimal values of the registration and simulation parameters, in terms of achieving in a minimum number of iterations the best trade-off between similarity of the images and smoothness of the transformation. These optimal values minimize the joint energy functional defined in a variational framework. In our experiments, we focus on the specific formulation of diffusion registration, but the exposed methodology can be directly applied to other non parametric registration schemes. The proposed method is validated over different scenarios, that include real medical images registration and non-rigid matching.

Important Links:

Go Back