# dicom files

**URL:** <https://discourse.itk.org/t/dicom-files/4999>\
**Category:** Beginner Questions\
**Tags:** dicom, threads\
**Created:** [May 20, 2022, 5:44am UTC](https://discourse.itk.org/t/dicom-files/4999 "2022-05-20T05:44:33Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![lang](https://discourse.itk.org/letter_avatar_proxy/v4/letter/l/0ea827/32.png) [@lang](https://discourse.itk.org/u/lang)\
**Post date:** [May 20, 2022, 5:44am UTC](https://discourse.itk.org/t/dicom-files/4999/1 "2022-05-20T05:44:34Z")

</div>

Multiple threads read dicom files concurrently, always in gdCMdatasethelper.cxx

if( tag.IsGroupLength() )

{

assert( vr == VR::UL );

}

if( tag.IsPrivateCreator() )

{

assert( vr == VR::LO );

}

Stop at assertion, what is the cause (ITK 4.13)

---

<div class="post-metadata">

**Author:** ![dzenanz](https://discourse.itk.org/user_avatar/discourse.itk.org/dzenanz/32/1093_2.png) [@dzenanz](https://discourse.itk.org/u/dzenanz)\
**Post date:** [May 23, 2022, 11:32am UTC](https://discourse.itk.org/t/dicom-files/4999/2 "2022-05-23T11:32:25Z")

</div>

If the read is single-threaded, does the same error occur? Can you share the offending image?

---

<div class="post-metadata">

**Author:** ![lang](https://discourse.itk.org/letter_avatar_proxy/v4/letter/l/0ea827/32.png) [@lang](https://discourse.itk.org/u/lang)\
**Post date:** [May 24, 2022, 2:55am UTC](https://discourse.itk.org/t/dicom-files/4999/3 "2022-05-24T02:55:59Z")

</div>

if the read is single-threaded，No errors，The image file that fails every time it is read is not fixed, the key is to compile the Release version without cras  
[IM293](https://discourse.itk.org/uploads/short-url/hKyZcvGuTpjl29hOZ2JLuv6dIku) (515.7 KB)  
hing, because release does not check Assert

---

<div class="post-metadata">

**Author:** ![dzenanz](https://discourse.itk.org/user_avatar/discourse.itk.org/dzenanz/32/1093_2.png) [@dzenanz](https://discourse.itk.org/u/dzenanz)\
**Post date:** [May 24, 2022, 7:54am UTC](https://discourse.itk.org/t/dicom-files/4999/4 "2022-05-24T07:54:16Z")

</div>

@mihail.isakov and @mathieu.malaterre might have some suggestions.

---

<div class="post-metadata">

**Author:** ![mihail.isakov](https://discourse.itk.org/letter_avatar_proxy/v4/letter/m/c4cdca/32.png) [@mihail.isakov](https://discourse.itk.org/u/mihail.isakov)\
**Post date:** [May 30, 2022, 1:46am UTC](https://discourse.itk.org/t/dicom-files/4999/5 "2022-05-30T01:46:12Z")

</div>

> [@lang](#):
>
> if the read is single-threaded，No errors

> [@lang](#):
>
> ITK 4.13

4.13 is already several years old. If you can reproduce the issue with the recent version and maybe a minimal example, try to open an issue here: [GDCM](https://sourceforge.net/projects/gdcm/) or write to the GDCM mailing list.

The asserts are related to CP-246, it is about datasets converted from implicit transfer syntax to explicit. It allows VR::UN, but never for _Group Length_ (retired) and _Private Creator_. I guess it is the reason for the check. Yes, it is only for _debug_ mode. The attached file has no issues, BTW.
