[Insight-users] information centric similarity measures

[David R. Haynor]

hi anja,

you don't say much about the exact application, but a couple of possibilties suggest themselves.  first, MI is best for cross-modality registration; if you have same-modality images, some form of correlation often works about as well, and you can certainly mix image matching criteria (appropriately scaled).  second, particularly for rigid-body registration, one way to handle things is to set up a different transformation for each pair of images and then enforce a consistency constraint by imposing penalties; then you are doing pairwise matching, but not n-way matching.

-dh

On Fri, 3 Dec 2010, Anja Ende wrote:

> Hello everyone,
>
> This is more of a general medical image registration question. Does
> anyone know of any literature or work being done on multi-modal
> similarity measures like MI or NMI but which are a bit more compact.
> One thing that I am trying to do is use multivariate mutual
> information and because of many images, my histogram size gets quite
> big. So, I am trying to look into other similarity measures which
> might be more friendly in this aspect.
>
> Thanks,
> Anja
> _____________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.html
>
> Kitware offers ITK Training Courses, for more information visit:
> http://www.kitware.com/products/protraining.html
>
> Please keep messages on-topic and check the ITK FAQ at:
> http://www.itk.org/Wiki/ITK_FAQ
>
> Follow this link to subscribe/unsubscribe:
> http://www.itk.org/mailman/listinfo/insight-users
>