[Insight-users] registering coronal & axial aquisitions?

[Luis Ibanez]

Hi Dean,

I'm not sure I understand what you want to do...
 From the coronal + axial data sets you want to
produce a single data set ?

Do you want to resample one of them to fit on
the coordinate system of the other ?

If registration is the goal, you may want to
try the following registration components:

1)  MeanSquaresImageMetric
2)  AffineTransform
3)  RegularStepGradientDescent
4)  NearestNeighborInterpolator.


(1) is because your images are of the same
     modality

(2) ... well, it is up to you... QuaternionRigid3D
     could also be an option...

(3) Is a stable optimier that works well with
     the MeanSquaresMetric as cost function

(4) It is the fastest interpolator...


The final output of a registration process is
a transform that you may use in the Resample
ImageFilter in order to reformat the moving
image on the coordinate system of the fixed
image.

You may want to look at the examples in the
Directory:

        Insight/Examples/Registration

That are associated to the SoftwareGuide.pdf
http://www.itk.org/ItkSoftwareGuide.pdf
Chapter 7 (pdf-page 167...)


Please let us know if you find any difficulties


   Thanks


     Luis


-------------------------------------
Dean Inglis wrote:
> Hi,
> 
> I have received two MR data sets: one coronal 
> acquisition and one axial, both are 512x512x22
> and of the same pulse sequence type.  Can
> someone provide some hints as to what ITK classes
> would be suited to constructing a single data set
> (not necessarily isotropic).  Is the registration
> framework ideal for this task?
> 
> regards,
> Dean
> 
> 
> _______________________________________________
> Insight-users mailing list
> Insight-users@public.kitware.com
> http://public.kitware.com/mailman/listinfo/insight-users
>