[ITK-users] [ITK] ResampleDICOM and Assertion failed: (slope == (int)slope)

[Matt McCormick]

Hi Nick,

A reasonable way move forward;

1) A patch could be made to ITK's internal GDCM to support non-integer slope's.
2) This patch could then be submitted to upstream GDCM.

Thanks,
Matt

On Wed, Sep 3, 2014 at 11:51 AM, Nick Patterson
<pattersonnp.work at gmail.com> wrote:
> Confirmed… a roll back to ITK 4.5.2 gets rid of this problem for the time being. It seems to reappear in any version after 4.6.
>
> Nick.
>
>
>>> Hi Nick,
>>>
>>> This issue may be coming up now because a bug was fixed before ITK 4.6
>>> which makes sure the rescale slope is applied when using GDCMImageIO
>>> for writing [1].
>>>
>>> HTH,
>>> Matt
>>>
>>>
>>> [1] http://itk.org/gitweb?p=ITK.git;a=commit;h=847ec2b79ac2cedaab6714678c0677793aa23a1e
>>>
>>> On Wed, Sep 3, 2014 at 11:10 AM, Nick Patterson
>>> <pattersonnp.work at gmail.com> wrote:
>>>> Hi,
>>>>
>>>> This is a problem that I have had before and have reported it as a ticket
>>>> on the GDCM Git page. I have written some code which is heavily dependent on
>>>> being able to read in a series of DICOM images, in this case they are of
>>>> PET/NM modality.
>>>>
>>>> When a series of images are provided, which have "non-integer" rescale
>>>> slope, the following error is returned at the final stage when
>>>> SeriesWriter->Update() is invoked.
>>>>
>>>> Assertion failed: (slope == (int)slope), function InverseRescaleFunction,
>>>> file /opt/SOURCE/GDCM/2.5/Source/MediaStorageAndFileFormat/gdcmRescaler.cxx,
>>>> line 67.
>>>> Abort trap: 6
>>>>
>>>>
>>>> I cannot provide my own code here, as it may make the problem less clear.
>>>> However, the problem is directly recreated using the ResampleDICOM.cxx
>>>> example from the ITK examples page. If a series of PT/NM images are run
>>>> through the code the above problem occurs. This is particularly problematic
>>>> since NM/PT modality images do often have non-integer rescale slopes and
>>>> currently I am not able to deal with images of this modality.
>>>>
>>>> There are 2 things to note:-
>>>> 1) I have reported this error with GDCM developer and this problem appeared
>>>> to have been resolved. I have previously been resampling NM/PT DICOM images
>>>> using the ResampleDICOM.cxx example in the past without issue. If I recall
>>>> correctly, ITK 4.5.2 worked with the latest GDCM release.
>>>>
>>>> 2) CT images (where the slope/intercepts are integer) work fine.
>>>>
>>>>
>>>> I shall re-log this on the GDCM pages, but since I am unsure that something
>>>> didn’t change in latest versions of ITK I wanted to report it here too.
>>>>
>>>> I would be much obliged if anyone would be able to test the
>>>> ResampleDICOM.cxx example with DICOM images which have a rescale slope which
>>>> is not-integer - and see if the problem can be recreated.
>>>>
>>>> I am currently using:- ITK 4.7 (from git) and pulled the latest GDCM from
>>>> Git Repo yesterday.
>>>>
>>>>
>>>> Regards, Nick.
>>>>
>>>>
>>>>
>>>> _____________________________________
>>>> Powered by www.kitware.com
>>>>
>>>> Visit other Kitware open-source projects at
>>>> http://www.kitware.com/opensource/opensource.html
>>>>
>>>> Kitware offers ITK Training Courses, for more information visit:
>>>> http://www.kitware.com/products/protraining.php
>>>>
>>>> Please keep messages on-topic and check the ITK FAQ at:
>>>> http://www.itk.org/Wiki/ITK_FAQ
>>>>
>>>> Follow this link to subscribe/unsubscribe:
>>>> http://public.kitware.com/mailman/listinfo/insight-users
>>>>
>>>> _______________________________________________
>>>> Community mailing list
>>>> Community at itk.org
>>>> http://public.kitware.com/mailman/listinfo/community
>>>>
>>
>
> _____________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.html
>
> Kitware offers ITK Training Courses, for more information visit:
> http://www.kitware.com/products/protraining.php
>
> Please keep messages on-topic and check the ITK FAQ at:
> http://www.itk.org/Wiki/ITK_FAQ
>
> Follow this link to subscribe/unsubscribe:
> http://public.kitware.com/mailman/listinfo/insight-users