You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Thanks for all your work, rt-utils is a great tool !
I am working with a MR image and sometimes, depending of the aquisitiion we need to apply a rotation so that the images are "axialized".
The result is that the pixelSpacing has a different value for x and y and the rtstruct generated is incorrect.
Here is a python script that allows to create a SEG and a RT
Please set force=True in rt_utils when loading the dcm with pydicom, I had to create dcm file to have a self content script
Thanks for opening this issue and for the detailed description.
I think your issue could be fixed if the 'synthetic' DICOM that you create would store correctly the PixelSpacing of the swapped axes. That should remove the need to have to modify the source-code within the image_helper module, wouldn't it?
Please let me know if doing something like that would also work.
Regarding the force=True of pydicom:
On one hand, I think it is safer have it as False in the tool, as we ensure dicom files are compliant and therefore we know what to expect within the tool, maximizing the chance of it working as intended.
On the other hand, I acknowledge that often in research we had to do some hacks to get newer stuff to work. So I also understand the need for more flexibility.
This is what I propose: I'll work on a way to implement an optional flag to allow the user to import non-compliant dicom files, with the warning that the tool might not work as intended if used like that. Then I'll bring that into the next release.
If you would like to work on that solution and create a PR, that would be fantastic as well and greatly appreciated!
Hi all,
Thanks for all your work, rt-utils is a great tool !
I am working with a MR image and sometimes, depending of the aquisitiion we need to apply a rotation so that the images are "axialized".
The result is that the pixelSpacing has a different value for x and y and the rtstruct generated is incorrect.
Here is a python script that allows to create a SEG and a RT
Please set
force=True
in rt_utils when loading the dcm with pydicom, I had to create dcm file to have a self content scriptwith MITK:
with MITK:
The text was updated successfully, but these errors were encountered: