Hi Frederico,
for the clarification of "Reinit" and Global Reinit":
- "global reinit" makes the first/red render-window display
transversal slices,
also if the image is not aligned along the MITK world coordinate system.
I.e. the global variant calculates a smallest bounding box in world
coordinates and makes this bounding box the render-window-world
- "reinit" takes the coordinate system of the selected mitk::Image as
the
render-window-world
There is also a bug for this since this caused some confusion among us either:
http://bugs.mitk.org/show_bug.cgi?id=8350
In this bug the problem with the MRI is mentioned too. I assume this behavior
only occurs with MRIs, doesn't it?
Best regards
Andreas
From: Federico Milano [mailto:[email protected]]
Sent: Freitag, 12. April 2013 03:11
To: [email protected]
Subject: [mitk-users] Reinit
Hi. I was wondering if somebody could kindly provide us with a general
explanation of what "ReInit" is doing behind the scenes.
This question arises since we are trying to use the ExtractSliceFilter, but we
are witnessing some puzzling results.
For example, we load a 3D volume (MRI/CT). The anatomical views are correct in
the mitkWorkbench. We execute ExtractSliceFilter (in a plugin). The first
problem is that the plane we are building to get a Geometry2D is initialized
with the "Frontal" orientation, but it is producing an "Axial" slice. After
experimenting for a while (maybe this is really basic), we have noticed a
second issue: that if you get the vtk image from the mitk image and then
reconvert it to a mitk image but without adding it the geometry of the original
mitk image, it appears a rotation in the generated mitk image, and the
anatomical views are wrong, but the PlaneGeometry is correct. Paradoxically, if
we do a reinit on the original mitk image, the same happens in the
mitkWorkbench, the image is rotated in one of the axis and the views don't
correspond to the anatomy anymore, but, in this case, the PlaneGeometry is
still incorrect, that is, it does not extract the correct slice when used with
the filter.
In conclusion, PlaneGeometry (and its use in ExtractSliceFilter) is consistent
with the vtk image, that is, the "frontal" plane is correctly located if we
convert the image to vtk and back to mitk without adding the original geometry,
but the problem is that, that image, is not oriented in an anatomical correct
way, so the views presented in mitkWorkbech are incorrect.
I find this kind of hard to express verbally, so please ask my for any
additional data you may need.
Thanks in advance,
Federico
------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
mitk-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mitk-users
