[Insight-users] Deformable Registration Final Parameters

[Torsten Rohlfing]

Hi Kevin:

In general, the "Moving" image is the one that should be reformatted, 
and afterwards it should resemble the (unchanged) "Fixed" image.

As for your concrete problem, I have the following questions:

1. DeformableRegistration7.cxx computed a 3D B-spline deformation, but 
BSplineWarping1.cxx applies a 2D transformation. Did you correct for that?

2. How do you get the B-spline coefficients from the first program to 
the second? It doesn't seem like DeformableRegistration7.cxx in its 
standard form supplies the coefficients in the form expected by 
BSplineWarping1.cxx? Did you add code to achieve that, or are you just 
reading the "deformationField" output file of 
DeformableRegistration7.cxx into BSplineWarping1.cxx? In that case, as 
they say on MythBusters, "well there's your problem" ;)

Best,
  Torsten

> Hello,
>
> What do the 'final parameters' in a b-splines deformable registration
> represent (I'm using DeformableRegistration7.cxx, revision 1.8)?  I was
> under the impression that the if I were to apply those parameters to an
> image via a b-splines warping algorithm (in this case,
> BSplineWarping1.cxx,revision
> 1.11) I would be be able to get an output image from the registration.  But
> this was not the case for me.
>
> That is, if I were to register image 'Fixed' to image 'Moving' using
> DeformableRegistration7.cxx, and using the final parameters from that
> registration on image 'Fixed' via BSplineWarping1.cxx, I'd be able to get an
> image that looks approximately similar to image 'Moving'.  Or does it work
> the reverse order?  But actually I've tried both ways and was unable to
> obtain what I was expecting.
>
>
> Thank you,
> Kevin

-- 
Torsten Rohlfing, PhD          SRI International, Neuroscience Program
 Research Scientist             333 Ravenswood Ave, Menlo Park, CA 94025
  Phone: ++1 (650) 859-3379      Fax: ++1 (650) 859-2743
   torsten at synapse.sri.com        http://www.stanford.edu/~rohlfing/

     "Though this be madness, yet there is a method in't"