[Insight-users] Using deformation fields when testing registration quality?

[Wes Turner]

At this point you undoubtedly have more than you need in answer to your
original question, but here is one more suggestion.  Instead of comparing
the deformation fields directly you can apply both to the voxels of the
image and evaluate the distance between the resulting positions.  Note that
for "dense" deformation fields this will be the same as comparing the
deformation fields directly.

- Wes

On Tue, Oct 27, 2009 at 12:43 AM, Karthik Krishnan <
karthik.krishnan at kitware.com> wrote:

> You could also try metrics to compare deformation fields.
>
> McCane, B., Novins, K., Crannitch, D. and Galvin, B., On Benchmarking
> Optical Flow, Computer Vision and Image Understanding, 84(1), pp
> 126-143, 2001
>
> See Section 5 titled Error Metrics in [1]. It provides 2 metrics to
> compare deformation fields. It also surveys other metrics used to
> compare deformation fields in literature and illustrates their
> shortcomings.
>
> Their methods are openly available at
> http://sourceforge.net/projects/of-eval/files/ .
>
> Motes : if you port them to ITK, and contribute it to IJ, I'd he happy
> to include them in the toolkit.
>
> Thanks
> --
> karthik
>
> On Mon, Oct 26, 2009 at 1:32 PM, Luis Ibanez <luis.ibanez at kitware.com>
> wrote:
> > Hi Motes,
> >
> > I would suggest that you subtract the Known deformation field
> > from the deformation field that results from the registration. This
> > will give you a "residual" deformation field. Then you could
> > compute the Average Magnitude of all the vectors in these
> > residual deformation field.
> >
> > It may also be worth to report the "largest magnitude" of that
> > residual deformation field.
> >
> > For subtracting two deformation fields, the itkSubtractImageFilter
> > should do the trick (although I have not tried this filter with images
> > of vectors).
> >
> > The computation of the average and maximum magnitude may
> > require you to write a customized while-loop with Iterators.
> >
> > It should take about 30~40   lines of code....  :-)
> >
> >
> > Please let us know if you find any problem,
> >
> >
> >     Thanks
> >
> >
> >         Luis
> >
> >
> >
> -----------------------------------------------------------------------------------------
> > On Mon, Oct 26, 2009 at 12:11 PM, motes motes <mort.motes at gmail.com>
> wrote:
> >> I would like to measure the quality of my modfied BSplineDeformable
> >> transform. I have used one of the ITK examples to create a known
> >> deformation of the fixed image that I use as the moving image.
> >>
> >> But how do I use this known deformation in the quality measure? Since
> >> the deformation field is applied to the fixed image and the result is
> >> the moving image I don't see what benefit it does when measuring the
> >> quality of the registration?
> >>
> >> How are deformation fields used in this context?
> >> _____________________________________
> >> Powered by www.kitware.com
> >>
> >> Visit other Kitware open-source projects at
> >> http://www.kitware.com/opensource/opensource.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
> >>
> > _____________________________________
> > 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
> >
> _____________________________________
> 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
>



-- 
Wesley D. Turner, Ph.D.
Kitware, Inc.
Technical Leader
28 Corporate Drive
Clifton Park, NY 12065-8662
Phone: 518-881-4920
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20091027/b6636371/attachment-0001.htm>