[Insight-users] All zeros displacement field when using DeformableRegistration1

[Luis Ibanez]

Hi Barbara,

1) One possibility is that there may be a mismatch in
    the magnitude of the displacement vectors relative
    to the pixel spacing in the image.


    1.1)   Could you please post your code to the list
    1.2)   Please indicate the pixel spacing of your image
    1.3)   Please indicate the max values of the deformation field


    You may want to try setting the DefaultPixelValue() to something
    different from zero, in order to verify if what you are seeing
    is your image being mapped outside of the field of view or
    if there is an intensity corruption on the resampling.


2) FEM is fine for what you are doing.
    You could try BSplines too, both techniques are in principle ok
    for this application.

    Note that if the Vertebrae are really close in shape you may
    also want to try the Demons deformable registration and the
    LevelSet deformable registration.




Please let us know about the questions above,


   Thanks


     Luis



----------------------
Barbara Garita wrote:

> Hello All:
>  
> I will like some advice concerning two issues about non-rigid registration:
> 1) I am using DeformableRegistration to register two 3D images (data 
> sets of human vertebrae). After I made the necessary accommodation for 
> 3D to the parameterlist and the code, I ran the DeformableRegistration1, 
> I checked for the correctness of the vectordeformationfield scalar 
> components  by extracting them and analyzing their values.  However, 
> when I use the Warperfilter and the vectordeformationfield to get back 
> to a scalar image I get a image of the right size but it is all zeros. 
>  
> 2) I am trying to non-rigidly register two image data set of human 
> vertebrae (I am interested in having a displacement field that gives me 
> correspondence between the surfaces (edges, contours...) of both 
> vertebrae, I am not really interested in the information inside the 
> vertebrae--), should I use another filter, like affine or B spline....
>  
> Thanks for you advice, Barbara
>  
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org
> http://www.itk.org/mailman/listinfo/insight-users