[Insight-users] Manual selection of control points

[Luis Ibanez]

Hi Ganesh,

1)  We don't have code for automatic selection of control points.

2) If you are looking for landmarks that can be used for initializing
    the Affine transform, you may have to select them manually and
    the use the LandmarkBasedTransfromInitializer.

    People in the list  may be able to recommend method for selecting
    salient features from images....

3) For the BSpline transform, we usually generate a regular grid,
    so there is no real need for automatic selection of points.

    You could use the class:

                   itkBSplineDeformableTransformInitializer

    that you will find in Insight/Code/Review

    its usage is described in the following Insight Journal paper:

    "Helper class for initializing the grid parameters of a
     BSpline deformable transform by using an image as reference"
     http://www.insight-journal.org/browse/publication/216
     http://hdl.handle.net/1926/1338


4) Regarding validation of deformable registration you may
    find interesting this other paper from the Insight Journal:


     "A Framework for Comparison and Evaluation of
      Nonlinear Intra-Subject Image Registration Algorithms"
      by Urschler M., Kluckner S., Bischof H.
     Institute for Computer Graphics and Vision,
     Graz University of Technology, Austria
     http://www.insight-journal.org/browse/publication/166
     http://hdl.handle.net/1926/561


5)  If you want to use fiducial marks as the basis for the
     evaluation, then you may find useful the following
     classes from the Image Guided Surgery Toolkit (IGSTK)
     (http://www.igstk.org):

      IGSTK/Source/
            igstkLandmark3DRegistrationErrorEstimator.cxx
            igstkLandmark3DRegistrationErrorEstimator.h
            igstkLandmark3DRegistration.h
            igstkLandmark3DRegistration.cxx

     They implement the method described in

     "Fiducial Point Placement and the Accuracy of
       Point-based, Rigid Body Registration",
         J.B West et al,
      Neurosurgery, pp 810-816, Vol.48, No.4, April 2001.



    Regards,


            Luis



--------------------------------------------------------------
On Tue, Mar 16, 2010 at 3:48 PM, Ganesh Narayanasamy
<nganesh76 at hotmail.com> wrote:
>
> Hi,
>  I am working on 3D-CT image registration problem. I start with performing
> affine registration on coarse images followed by a non-rigid registration
> (Demons or B-spline) on fine images.
>
> a) Is there a code for manual selection of control points that could guide
> the affine and then the non-rigid registration module?
>
> b) Once the registration is done, I would like to evaluate the registration
> error using a different non-rigid registration technique (say, Thin plate
> spline). This would need another set of manually selected fiducial markers
> on the fixed and the registered image volume. My question is: Is it possible
> to apply the non-rigid transformation parameters on these fiducial markers
> and get its corresponding coordinates in the new image space?
>
> Thanks,
> Ganesh
>
> ************************************************************************
> "What the mind of man can conceive and believe, It can achieve" - Napolean
> Hill
> -Ganesh Narayanasamy,Ph.D., PostDoc-University of Kentucky Cancer Center
> Earlier work: http://www.sciencedaily.com/releases/2007/05/070504133017.htm
> ************************************************************************
>
> ________________________________
> Hotmail: Trusted email with powerful SPAM protection. Sign up now.
> _____________________________________
> 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
>
>