Retrieving DVH Sequence in RT Dose IOD

[mlundemann]

Hi,

I've set up a DICOM Q/R service at my research institution, using dcmqrscp, findscu and movescu. I would like to retrieve DVH Sequence Data from the database connected to our Treatment Planning Software (TPS). At the moment I'm able to retrieve RT Dose files containing pixel data, but how do I specify (preferably at the command line) that the DVH Sequence should be included?

Thanks,
Michael

[J. Riesmeier]

Are you talking about returning DVH Sequence (3004,0050) with a C-FIND response message from the Q/R SCP?

This attribute is not supported by the dcmqrscp tool. Are you sure that it is available from the TPS database?

[mlundemann]

Dear J. Reismeier,

Thank you for your response.

Yes! It's exactly what I want to do, so I guess there's no other way than to modify the dcmqrscp. DVH data has been calculated by the TPS and is exported by default, when using the TPS UI, so I believe it's only a matter of a correct C-FIND response.

/Michael

[J. Riesmeier]

You cannot expect that all (most) DICOM attributes are supported by an Q/R SCP. Why don't you just retrieve (i.e. download) the RT Dose object and the access the DVH Sequence in that object (if available)? I guess you are doing the same for Pixel Data, which contains the actual dose information.

[mlundemann]

Yeah, but my problem is that the DVH Sequence isn't included (available) in the RT Dose Object that I retrieve via dcmqrscp. However, the DVH data is estimated and available in the TPS and when I, on the contrary, manually export the same RT Dose Object from the TPS (via the user interface), the DVH Sequence is included.

So I thought I could "force" dcmqrscp to include the DVH Sequence in the RT Dose Object, by using the query key (3004,0050) in movescu, but as you say that option is (obviously) not supported by the Q/R SCP.

I have 400 patients that I wish to analyze, which makes a manual export impossible.

My work-around is to re-estimate the DVH from the Pixel Data and the RT Structure Set in MATLAB, but it's really a duplication of effort and it induces minor differences due to partial volume effects, when compared to the DVH estimated by the TPS.

[J. Riesmeier]

However, the DVH data is estimated and available in the TPS and when I, on the contrary, manually export the same RT Dose Object from the TPS (via the user interface), the DVH Sequence is included.

That's really strange, but this should have nothing to do with the way the Sequence is requested, i.e. via C-FIND or not. Maybe, there is a configuration option in the TPS or...

[Michael Onken]

Hi,

just to be 100% sure , you can use storescp instead of dcmqrscp as a receiver for the data. You can add storescp's option "--bit-preserving" on top to receive every single bit as it was sent over the line.

Best,
Michael