questions > problem with converted file name
Showing 1-5 of 5 posts
Aug 8, 2016 03:08 PM | Carlo Lucci
problem with converted file name
Hi everybody,
I am a student and I am doing a research project on a big number of patients. I encountered an issue with the file output name. After I convert the files, I need to know which file contains which sequence but the program has just the %s command which gives me the sequence number (e.g. 601). Is it possible someway to have the SeriesDescription (e.g. T1 SE) inside the output name?
Thanks to anybody who can help me!
Carlo
I am a student and I am doing a research project on a big number of patients. I encountered an issue with the file output name. After I convert the files, I need to know which file contains which sequence but the program has just the %s command which gives me the sequence number (e.g. 601). Is it possible someway to have the SeriesDescription (e.g. T1 SE) inside the output name?
Thanks to anybody who can help me!
Carlo
Aug 8, 2016 04:08 PM | Chris Rorden
RE: problem with converted file name
Carlo-
I think you want to include the "%p" option for protocol name
https://www.nitrc.org/plugins/mwiki/index.php/dcm2nii:MainPage#General_Usage
You can see the results of a command name interactively with MRIcroGL: Choose the Import/Convert command and put "%p_%f_%t_%s" and you will see the text appear "Example output filename: '/MPRAGE__myFolder_19770703150928_1.nii.gz', showing that we are describing the protocol (%p), the folder name (%f), the session time (%t) and the series number (%s).
I think you want to include the "%p" option for protocol name
https://www.nitrc.org/plugins/mwiki/index.php/dcm2nii:MainPage#General_Usage
You can see the results of a command name interactively with MRIcroGL: Choose the Import/Convert command and put "%p_%f_%t_%s" and you will see the text appear "Example output filename: '/MPRAGE__myFolder_19770703150928_1.nii.gz', showing that we are describing the protocol (%p), the folder name (%f), the session time (%t) and the series number (%s).
Aug 8, 2016 05:08 PM | Carlo Lucci
RE: problem with converted file name
Dear Chris,
many thanks for your early answer! I tried with the %p option for protocol name but since the data are anonymised it just put "anonymus" in that part of the output name. I looked in the metadata of the DICOM images that the name of the sequences is in the SeriesDescription but using the program with the %s command it collects the information from SeriesNumber. That is why I was trying to make another command for the output name which could collect the data from the SeriesDescription part of the file. I hope it makes sense for you
many thanks for your early answer! I tried with the %p option for protocol name but since the data are anonymised it just put "anonymus" in that part of the output name. I looked in the metadata of the DICOM images that the name of the sequences is in the SeriesDescription but using the program with the %s command it collects the information from SeriesNumber. That is why I was trying to make another command for the output name which could collect the data from the SeriesDescription part of the file. I hope it makes sense for you
Aug 9, 2016 08:08 PM | Chris Rorden
RE: problem with converted file name
Carlo-
I really think this is a problem with your anonymizer rather than my software. The %p option fills in the "Protocol Name" (0018,1030) and will fall back to Series Description (0008,103E) or Sequence Name (0018,0024) only if the Protocol Name is empty. I do not think Protocol Name should be anonymized, and if it is, why not remove it rather than insert a bogus value?
In any case, I have created a new version you can test out (9Aug2016)
https://github.com/neurolabusc/dcm2niix
that includes two new options:
%d Description: Series description (0008,103E) [NEW]
%z Sequence Name (0018,0024) [NEW]
Feel free to compile this version and try it out on your system.
I really think this is a problem with your anonymizer rather than my software. The %p option fills in the "Protocol Name" (0018,1030) and will fall back to Series Description (0008,103E) or Sequence Name (0018,0024) only if the Protocol Name is empty. I do not think Protocol Name should be anonymized, and if it is, why not remove it rather than insert a bogus value?
In any case, I have created a new version you can test out (9Aug2016)
https://github.com/neurolabusc/dcm2niix
that includes two new options:
%d Description: Series description (0008,103E) [NEW]
%z Sequence Name (0018,0024) [NEW]
Feel free to compile this version and try it out on your system.
Aug 10, 2016 08:08 AM | Carlo Lucci
RE: problem with converted file name
Dear Chris,
thank you so much! I used the anonymizing option of the DICOM viewer from my hospital but unfortunately I didn't know it would have anonymized also the protocol name information. Since we are working on a big number of patients I should have had to download again from the archive without anonymization all the images, but thankfully your new version worked perfectly with the images I already had.
Many thanks again and best regards,
Carlo
thank you so much! I used the anonymizing option of the DICOM viewer from my hospital but unfortunately I didn't know it would have anonymized also the protocol name information. Since we are working on a big number of patients I should have had to download again from the archive without anonymization all the images, but thankfully your new version worked perfectly with the images I already had.
Many thanks again and best regards,
Carlo