[vtkusers] Yet another VTK, DICOM and Orientation problem

[Jean-Christophe Fillion-Robin]

Hi Romain,

To learn which rotation matrix would be needed, you could load the dataset
into 3d Slicer and inspect the direction cosine in the Volumes module. See
https://download.slicer.org/

That said, considering that VTK is orientation-less (there are no direction
cosines associated with vtkImageData), building an medical image viewer is
a significant amount of work.

Also worth noting that there is a proposal to add orientation into
vtkImageData but some more validation is needed, see
https://www.slicer.org/wiki/Documentation/Labs/VTK-Orientation

Hth
Jc

On Tue, Dec 11, 2018 at 8:12 AM Romain LEGUAY <romain.leguay at gmail.com>
wrote:

> Hello everyone,
>
> I'm trying to create a viewer of dicom with anatomical views (axial,
> sagittal and coronal) from a radiologist point of view.
>
> As starting point, I use the example provided in vtkDICOM module :
> testDicomDisplay
> <https://github.com/dgobbi/vtk-dicom/blob/master/Examples/TestDICOMDisplay.cxx>
> .
> It uses vtkDICOMReader with SetMemoryRowOrderToFileNative activated.
>
> When I try to display an axial CT scan, I don't obtain expected
> orientation as you can see in pic1: the sagittal view is in Posterior ->
> Anterior direction.
> [image: pic1.png]
>
> When I try to display a sagittal CT scan, I have different orientations:
> [image: pic2.png]
>
> To fix this orientation problem, I tried to create some transformation
> matrices that converts
> Patient Coordinates (DICOM) to World Coordinate (VTK coordinates).
>
> To do this, I follow those steps:
> * get the patient matrix (Pm) using the method
> vtkDICOMReader::GetPatientMatrix (if I understand correctly, this matrix is
> in DICOM space)
> * create the rotation (Rm) from DICOM space to VTK space (rotation on VTK
> X-axis around the origin point):
> 1 0 0 0
> 0 -1 0 0
> 0 0 -1 0
> 0 0 0 1
> * Multiply Pm by Rm to obtain the matrix PRm to go from Patient
> Coordinates to VTK Coordinates.
> * Use the Matrix PRm with a point in IJK coordinates (voxel coordinates)
> multiplied by the spacing.
>
> I know I made a mistake because I don't have the result expected but I
> don't know what I did wrong...
>
> Can someone provide me some tips please?
>
> Thank you,
>
> Romain
>
>
>
>
> _______________________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.html
>
> Please keep messages on-topic and check the VTK FAQ at:
> http://www.vtk.org/Wiki/VTK_FAQ
>
> Search the list archives at: http://markmail.org/search/?q=vtkusers
>
> Follow this link to subscribe/unsubscribe:
> https://public.kitware.com/mailman/listinfo/vtkusers
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://public.kitware.com/pipermail/vtkusers/attachments/20181211/e19ae568/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: pic1.png
Type: image/png
Size: 572778 bytes
Desc: not available
URL: <https://public.kitware.com/pipermail/vtkusers/attachments/20181211/e19ae568/attachment-0002.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: pic2.png
Type: image/png
Size: 544089 bytes
Desc: not available
URL: <https://public.kitware.com/pipermail/vtkusers/attachments/20181211/e19ae568/attachment-0003.png>