Steve Pieper

Hi @neurolabusc , sorry to leave you hanging on this but it's not really in my critical path and I know how much time it can take to iron out...

I haven't looked into this for a very long time either, but this code indicates that pyside should now be usable from inside a C++ application. https://github.com/pyside/pyside-setup/tree/dev/examples/scriptableapplication

Is there a version of the code in github or pypy that should work for the compute example?

Yes, that one works - thanks!

For reference, Slicer provides a [DICOMPatcher](https://slicer.readthedocs.io/en/latest/user_guide/module_dicompatcher.html) module that handles a number of common issues. This is another set of code (like the [plugins](https://www.slicer.org/wiki/Documentation/4.10/Modules/DICOM#DICOM_Loading.2FSaving_and_Plugins)) that we could factor out into `highdicom`...

Some cases will be very simple, but when you get into sophisticated imaging scenarios like diffusion MRI the task is very complex (see for example [dcm2niix](https://www.nitrc.org/plugins/mwiki/index.php/dcm2nii:MainPage) and related efforts like...

I would vote for always creating new SOPInstanceUIDs for any derived instance (the original UIDs can be referenced if someone needs to track them down). Also as a general rule...

> Are you arguing in favor of skipping all private attributes and not including them in the enhanced data sets at all? Yes, my preference would be that the normalized...

> Steve, but what about private tags that cannot be mapped into standard ones? I don't think it makes a lot of sense to convert instances that don't map to...

My main point is that these "non-mappable private tags" are essentially opaque to us, and in my experience they probably have very little meaning if they are removed from their...