It is properly a problem in ITK.
When i try to read the DCM file using ITKSimple, i get the same problem.

Screenshots and DCM Data

2D test image

Load the attached 2D image.

screenshot2

screenshot3

I have another image (standard direction) with spacing of 40, 40, 48 and also get strange results. See screenshot3. Here, the manually segmented layer is not fully included in the interpolation.

screenshot1

screenshot2

screenshot3

screenshot1

[a74f05]: Merge branch 'bug-17313-statemachineRenameCurveModelToMitralModel'

Merged commits:

2014-02-24 11:26:24 Bastian Graser [e4327f]
Renamed CurveModel to MitralModel

Maybe somehow connected to T15363 ?

Looked at mitkItkImageFileReader.cpp at line 79
after calling

[b296fe]: Merge branch 'bug-16508-StatemachineExtension'

Merged commits:

2013-11-26 14:52:34 Bastian Graser [a8d723]
Added Shift+RMB and Shift+LMB interactions and states to implement

New remote branch pushed: bug-16508-StatemachineExtension

screenshot

If you do it the other way around, MITK produces a memory leak:

Currently it seems that the Image-Geometry and the Segmentation-Geometry are checked to be equal! Otherwise there is a warning. Additionally, A warning is shown to indicate that a reinit is necessary, when the image-combobx has changed (?).

[393e6f]: Merge branch 'bug-16469-segmentationwarning'

Merged commits:

2013-11-27 14:59:41 Bastian Graser [18f136]
checking world-geometry with selected image and segmentation node geometry

screenshot

Apparently CMake was not the problem.
The problem is NSIS

Thanks michael!

i am having a look at this issue

The patch seems fine.

Has been fixed apparently.

New remote branch pushed: bug-16243-NewIndexToWorldFunction

[6f7fd1]: Merge branch 'bug-16243-NewIndexToWorldFunction'

Merged commits:

2014-01-15 19:19:39 Bastian Graser [f8539c]
added IndexToWorld method for itk::index and mitk::index3d

What is the current status here??

compiler output

Nightly Dartclients work!

The problem lies within visual studio.

i have a look

Crashes in QmitkDeformableClippingPlaneView.cpp
in line 381

Under windows the clipping plane plugin is always crashing when selecting an image in the datamanager and clicking the button "Create new clippping plane"

Bug only occurs, when segmentation plugin was activated before.

The bug appears when the segmentation view is started but deactivated (in the background).

New remote branch pushed: bug-16118-philipsdicom

In my opinion we should stay as close as possible to ITK.
In ITK there is no geometry in Pointsets.
Therefor i would change the points before saving instead of saving an additionaly geometry.

mit SetIndexToWorldTransform(..) setzt man eine Transformation (bestehend aus einer Matrix und einem Vector). Das Spacing ist in der Matrix eingrechnet, somit wird auch das Spacing überschrieben. Sofern kommt da normalerweise nicht dasselbe raus.

As the behavior is not unexpected, i close this bug.

In my opinion it would be more intuitive to separate the spacing information from the matrix.

I am very sorry for the late reply. Did not see that bug.

Hi Michael, whats the status?
is this solved?

[e74202]: Merge branch 'bug-15354-datastoragecombobox'

Merged commits:

2013-06-11 18:39:19 Bastian Graser [b82f25]
added secutiry check to prevent crash

New remote branch pushed: bug-15354-datastoragecombobox

I am also confronted with this task now.

todo: delete mitkErodeTool etc.

New remote branch pushed: bug-15170-morpho4

New remote branch pushed: bug-15170-morpho2

New remote branch pushed: bug-15170-morpho6

New remote branch pushed: bug-15170-morpho

Moved to new GUI .. everything done so far.
if no errors occur, this bug can be closed

New remote branch pushed: bug-15170-morpho3

Morphological Operators moved to new widget. Implementation outsourced into module. 4D support enabled. deleted depcreated morphological tools.

Morphological Operations were implemented

We assume that at Joesphs indicated position in the code, extractSliceFilter is called with an image geometry, though it is expecting a bounding geometry.

A solution would be NOT to limit the measurement Interactions by Image / World geometry bounds. I do not think there are any disadvantages in doing so.

Hash of tested Version: 04e09a17

Also, if image is a segmentation, measurement is not possible.

Thanks for the correction.
Yeah, apparently this was a bug.
The new behavior you describe is how it should be!

In the current source code, there is no call
ChangeImageGeometryConsideringOriginOffset(false)
in the FitGeometry method anymore.

i'll check it out

An extra slice is produced, which is initialized with the value "0". This is "grey" in CT images.

Apparently this is done un purpose:

Correction: Padding is not really done on purpose.