Exporting an RT Plan DICOM from matRad

[oapaza12]

Hello everyone,

I have a question regarding matRad. Is it possible to export the plan generated in matRad as a DICOM file, specifically in the RT Plan format?

Thank you for your assistance!

[wahln]

Hi,
it is principally possible on the current dev branch, at least for photons. Particles are not yet supported, but if it is required (by you?), we may look at it (or help you in implementing it).

The dicom export is implemented in matRad_DicomExporter.

Warning

We are a little bit hesitant with allowing RTPlan exports (a little bit anxious of unconscious misuse of generated RTPlans in clinical workflows). Thus on the branch the DICOM export of RTPlans is protected by two barriers:

1. Needs to be enabled in the dicom exporter object, e.g., dcmExporter.enableRtPlanExport = true;
2. An error throwing safeguard needs to be commented out in the corresponding function of the exporter to make sure users know what they are doing!

In the first example, the last lines perform an exemplary dicom export:

matRad/examples/matRad_example1_phantom.m

Lines 146 to 154 in b25659d

	%%
	% We export the the created phantom & dose as dicom. This is handled by the
	% class matRad_DicomExporter. When no arguments are given, the exporter searches
	% the workspace itself for matRad-structures. The output directory can be set by
	% the property dicomDir. While the different DICOM datasets (ct, RTStruct, etc)
	% can be exported individually, we call the wrapper to do all possible exports.
	dcmExport = matRad_DicomExporter();
	dcmExport.dicomDir = [matRad_cfg.primaryUserFolder filesep 'dicomExport'];
	dcmExport.matRad_exportDicom();

dicomDir can be every directory, in above example it uses the primary user folder (new concept) stored in the matRad_cfg object (which thus needs to be instantiated), but you can also use any absolute path to your liking.

[oapaza12]

Hello Niklas Wahl, thank you for your response. Fortunately, I'm working specifically with photons using the developer mode function on the TrueBeam Linac. I'll try to resolve the issue.

[edwardwang1]

Hi @oapaza12,

I am also trying to use matRad to export an RTPlan for a Varian TrueBeam. Are you using Eclipse? I think it could be very helpful for both of us to compare notes.

[oapaza12]

Hello EdwardWang1,
Absolutely, we can share our notes. Yes, I’m using the Eclipse TPS and executing XML files in developer mode on the TrueBeam linac.

[oapaza12]

This is my personal email: [email protected]. Please send me a message so we can discuss our projects in more detail.

[wahln]

It would also be really nice to create a bigger discussion topic here on GitHub on this. It seems like that recalculation together with creation of own datasets is less trivial with other machines and algorithms than with the one we used for validation, and there is probably big potential for improvement. I will create something.

[wahln]

Opened this thread for shared discussion on this topic:
#772