[Insight-users] DigitallyReconstructedRadiograph

[Luis Ibanez]

Hi Watermark,



      "Glimpsing at the source leaves no doubt"



    a) Open the file in a text editor

    b) Search for "-t", and you get to line 174
       where the translation values are assigned
       to "tx,ty,tz".

    c) Seartch for "tx", and you get to line 494
       where tx,ty,tz are assigned to the vector
       "translation"

    d) Search for "translation" and you get to
       line 501, where it is assigned to the
       "transform".

    e) Look at the type of "TransformType", and you
       get to line 486, where it is declared as

       typedef itk::CenteredEuler3DTransform< double >  TransformType;

    f) Search for "transform" and you get to line 609,
       where it is assigned to the ResampleImageFilter

    g) Open the ITK Software Guide

       http://www.itk.org/ItkSoftwareGuide.pdf

       search for ResampleImageFilter and you will
       find the section on "Geometrical Transformations".

    h) The guide explains that the transforms used
       by the ResampleImageFilter is the one that goes
       from the physical space of the output image to
       the physical space of the input image (the image
       from which the intensity pixel data is read).

       Please read this section with attention.

       Lack of reading this section is the source of
       your confusion.


    i) Therefore, the translations tx,ty,tz that you type
       in the command line option "-t", will be used to
       be added to the coordinates of a pixel from the
       output image (the DRR), to compute the coordinates
       of the pixels from which intensity values will be
       taken in order to reconstruct the X-Ray absortion.

       You can therefore describe the translations as
       "moving the camera" or the X-Ray machine that is
       being used for acquiring (synthetizing) the X-ray
       projection.

--


            Welcome to Open Source software.



    Regards,


       Luis



------------------
MedVision wrote:
> Hi all,
> 
>    I'm new in ITK and I tried DRR generation according to the DigitallyReconstructedRadiograph1.cxx example. 
>    My understanding is that the transform should be applied to the volume data. However in the example program it is noted: "<-t float float float>" as "translation parameter of the camera", I'm wondering whether it is the transformation for the volume or the camera.  As for the generated image, do I watch from the focalpoint facing the volume, or else ? 
> 
>    Thanks and Regards
> 
> watermark
> 
> 
> 
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org
> http://www.itk.org/mailman/listinfo/insight-users
>