[Vv] clitkDicomRTStruct2Image issues

Tue Jan 24 16:25:49 CET 2017

Thank you for your remarks on Docker. You are the first known user of 
the Docker image ! We will talk about it and maybe reconsider it to be a 
good first look for vv.
It's possible that some command line tool helps are not exactly updated. 
We start to migrate all codes on github. So you can report an issue on 
it:
https://github.com/open-vv/vv
(we will communicate on it latter).

Thank you
Thomas

On 2017-01-23 16:01, Jonathan Martens wrote:
> Hi Thomas,
> 
> Thanks for your prompt reply and pointing me in the right direction on
> how to use the command line tools. I already had the impression I was
> doing something the wrong way. After creating a .mhd file for my CT
> scan, I was able to extract structure bitmasks.
> 
> I appreciate your remark on the command line tools, but for me that is
> the only way to go now, as I need a quick way to create a proof of
> concept and we are most likely to run Elastix using SimpleElastix from
> Matlab in the end anyhow.
> 
> Perhaps it is worth considering also validating the command line tools
> in the Docker image as that is a very quick and easy way for people to
> start working with your programs.
> 
> One issue I already discovered is that not all tools print their
> version information correctly. Unfortunately, I do not have access to
> your issue tracker (yet) to send in a documented bug report.
> 
> Kind regards,
> 
> Jonathan
> 
> -----Original Message-----
> From: vv [mailto:vv-bounces at creatis.insa-lyon.fr] On Behalf Of Thomas 
> BAUDIER
> Sent: maandag 23 januari 2017 15:15
> To: vv at creatis.insa-lyon.fr
> Subject: Re: [Vv] clitkDicomRTStruct2Image issues
> 
> Hi Jonathan,
> 
> Thank you to report this problem.
> 
> I just want to be sure you use the clitkDicomRTStruct2Image tool
> correctly:
> The tool takes a RTStruct file in input (-i) with 1 or more structures
> and a CT image (-j). Be sure to have a complete CT image with all
> slices (I saw you just take a .dcm file, but I don't know if it's just
> a slice or not). I prefer to use an mhd image for the CT. You can
> convert a DicomSerie into a .mhd file with clitkDicom2Image -o
> output.mhd *.dcm And the outputs are each structures in a different
> .mhd file. The name of these files are VARnbStruct.mhd with Var the
> output -o and nbStruct the number of your structure.
> 
> The Docker image isn't very suitable for command line tools, we did
> for vv exclusively. So we didn't test the command line tools with
> Docker.
> The best is to use the git version to be sure to have the newest
> running version.
> 
> Thomas
> 
> On 2017-01-23 12:55, Jonathan Martens wrote:
>> Dear all,
>> 
>> I am trying to work with Elastix (http://elastix.isi.uu.nl) and
>> therefore need to convert structure sets from DICOM RT-STRUCT files to
>> the proper format accepted by Elastix. Elastix seems to require files
>> with a metaimage format in .mhd/.raw format.
>> 
>> On my search on the internet I found an older thread in this e-mail
>> list referencing a tool (part of vv) to convert RT-STRUCT files to the
>> required format
>> (https://www.creatis.insa-lyon.fr/pipermail/vv/2013-February/000285.html).
>> 
>> 
>> I got the Docker container running as suggested per
>> https://www.creatis.insa-lyon.fr/rio/vv?action=show&redirect=VV#line-1
>> 09
>> [1] on Windows 7 using the following:
>> 
>>   docker pull tbaudier/vv
>> 
>>   docker run -name vv -e DISPLAY= $(docker-machine ip):0 -v
>> $(pwd):/home tbaudier/vv
>> 
>> in combination with XMing running with the following command line (had
>> to add the -ac switch to the default command line)
>> 
>>   Xming.exe :0 -ac -clipboard -multiwindow
>> 
>> After that I am able to start the container and run vv in a X11
>> terminal on Windows 7. Besides I am able to attach a shell:
>> 
>>   docker exec -it vv /bin/bash
>> 
>> From that I can run the command line tools. But when I try to extract
>> data from my RT-STRUCT file like this:
>> 
>> p=P0162C0006I13345846;
>> 
>> rtstruct=RTSTRUCT/1.2.246.352.71.2.78480665410.4114624.20161221081018/
>> RTSTRUCT.1.2.246.352.71.4.78480665410.142535.20161221114758.dcm;
>> 
>> 
>> ct=CT/1.3.6.1.4.1.32722.162.6.55613550350070884234080940853841385982/C
>> T.1.3.6.1.4.1.32722.162.6.98938198109550082302924371538367888818.dcm;
>> 
>> 
>> d=/home/DICOM;
>> 
>> params=-v
>> 
>> clitkDicomRTStruct2Image $params \
>> 
>> -i $d/$p/$rtstruct \
>> 
>> -j $d/$p/$ct \
>> 
>> -o $d/$p/test.mhd
>> 
>> I seem to run into issues, as clitkDicomRTStruct2Image seems to,
>> consequently, return correct .mhd files, but .raw with a size of 256kB
>> that contain only NULL values (512 * 512 * 1 = 262144).
>> 
>> I would have expected to have some data in it as the RT-STRUCT file
>> does contain contours. Even if I extract a specific contour for all
>> slices of the CT set referenced from the RT-STRUCT I seem to get no
>> data in any file.
>> 
>> Am I doing something wrong? Am I expecting something wrong?
>> 
>> Kind regards,
>> 
>> Jonathan
>> 
>> Links:
>> ------
>> [1]
>> https://www.creatis.insa-lyon.fr/rio/vv?action=show&redirect=VV#li
>> ne-109 _______________________________________________
>> vv mailing list
>> vv at creatis.insa-lyon.fr
>> http://www.creatis.insa-lyon.fr/mailman/listinfo/vv
> _______________________________________________
> vv mailing list
> vv at creatis.insa-lyon.fr
> http://www.creatis.insa-lyon.fr/mailman/listinfo/vv
> 
> _______________________________________________
> vv mailing list
> vv at creatis.insa-lyon.fr
> http://www.creatis.insa-lyon.fr/mailman/listinfo/vv