[Insight-users] Narrow band to image registration

[Luis Ibanez]

Hi Eduard,


1) Yeap, coffee and systematic experimentation
    are the way to go. Make sure that you connect
    a Command/Observer and start using the simplest
    possible Transform (e.g. TranslationTransform).



2) Nope !
    The points of the Narrow band are not used
    at all by the KernelTransform.  The points
    in the NarrowBand are the "fixed object"
    participating in the registration.

    Please forget about the KernelTransform until
    you get familiar with the registration method
    using standard simple transforms such as
    Translation, Rigid and Affine.



Regards,


    Luis



------------------------
Eduard Schreibmann wrote:

> Quoting Luis Ibanez <luis.ibanez@kitware.com>:
> 
> 
>>Hi Eduard,
>>
>>
>>1) Won't be able to help you with the code
>>    that doesn't work... unless you tell us
>>    what happens when you run it:
>>
>>    - does it crash  ?
>>    - does it aborts ?
>>    - does it runs indefinitely ?
>>
>>    a bit more information may be useful  :-)
> 
> 
> None of the above, it works fine, but it is not able to find a translation
> that was artificially set in the moving image. Instead of 5,0,0 it finds a
> translation of 6,7,3 :) I suppose more cofee and different
> setings/optimizers will work ?
> 
> 
>>2)  The Kernel Splines simply use the point
>>     pairs in order to define Vectors, then
>>     interpolates those vectors in space.
>>     You can think of it just as the BSpline
>>     deformable transform: A mechanism for
>>     representing a deformation field.
>>
>>     More details on how this transforms work
>>     are available at:
>>http://www.itk.org/Insight/Doxygen/html/classitk_1_1KernelTransform.html
>>
>>     and of course in the original paper
>>
>>      IEEE TMI,  Vol. 16, No. 3 June 1997
>>      Davis, Khotanzad, Flamig, and Harms,
>>
> 
> 
>>From the documenation :
> 
> "KernelTransform: Provides support for defining source and target landmarks"
> 
> The target landmarks are the point set from the narrow band, but what will
> be the source landmarks ? I have one image as source. What do I miss
> out/confuse  ?
> 
> 
> 
>>3)  If you MRI image is hightly inhomogeneous, you
>>     should apply first the correction method in
> 
> 
> Did, it works nice.
> 
> Thank you for the replays.
> Edi
>