sorting dicom files into dirs - DICOM

This is a discussion on sorting dicom files into dirs - DICOM ; Hi, What would be "best" tool to sort a number of DICOM files into PatientID\SeriesID\StudyID\AcquisitionTime directory structure? (would be nice if I could change/specify the directory hierarchy) I could use storescp to resort studies according to UID, which has the ...

+ Reply to Thread
Results 1 to 6 of 6

Thread: sorting dicom files into dirs

  1. sorting dicom files into dirs

    Hi,

    What would be "best" tool to sort a number of DICOM files into
    PatientID\SeriesID\StudyID\AcquisitionTime directory structure?
    (would be nice if I could change/specify the directory hierarchy)

    I could use storescp to resort studies according to UID,
    which has the benefit of avoiding possible PatientID, SeriesID clashes
    (I have many studies from different places, and would not you know
    each is using different scheme for incrementing the IDs, so a clash is
    possible), but it does not feel like the right approach.

    Is there such a tool, or should I be writing my own ?

    Regards,
    Igor



  2. Re: sorting dicom files into dirs

    "Igor Okulist" wrote in message news:...
    > Hi,
    >
    > What would be "best" tool to sort a number of DICOM files into
    > PatientID\SeriesID\StudyID\AcquisitionTime directory structure?
    > (would be nice if I could change/specify the directory hierarchy)
    >
    > I could use storescp to resort studies according to UID,
    > which has the benefit of avoiding possible PatientID, SeriesID clashes
    > (I have many studies from different places, and would not you know
    > each is using different scheme for incrementing the IDs, so a clash is
    > possible), but it does not feel like the right approach.
    >
    > Is there such a tool, or should I be writing my own ?
    >
    > Regards,
    > Igor


    Hi Igor:

    You can use the simple_storage from ctn with the -n option, that uses a file
    where you decide which dicom tags form part of the path, subdirs and filename.
    I hope it hepls

    Pablo Sau
    http://cdmedicpacsweb.sourceforge.net

  3. Re: sorting dicom files into dirs

    "Igor Okulist" wrote in message news:...
    > Hi,
    >
    > What would be "best" tool to sort a number of DICOM files into
    > PatientID\SeriesID\StudyID\AcquisitionTime directory structure?
    > (would be nice if I could change/specify the directory hierarchy)
    >
    > I could use storescp to resort studies according to UID,
    > which has the benefit of avoiding possible PatientID, SeriesID clashes
    > (I have many studies from different places, and would not you know
    > each is using different scheme for incrementing the IDs, so a clash is
    > possible), but it does not feel like the right approach.
    >


    Why would you want to sort with series at a higher level than study
    id? If the data is compliant, there will only every be one entry under
    each series?

    More to the point, the modality you are using may be generating nice
    values for Series ID and Study ID, but many do not. DICOM doesn't
    mandate that either of them be unique. A reasonable interpretation of
    the semantics of the attributes would be to make Series Ids unique
    within the scope of the study and study ID unique within some period
    of time in which multiple studies are generated. Most vendors do a
    decent job with Study ID, when they bother. Series ID is another
    story. Very often I've seen a CT or MR study with 6 or 7 series, all
    labeled with the value "1" for the series ID. You can sort by the IDs,
    but for some manufactures, all your images are going to end up in the
    same directory.

  4. Re: sorting dicom files into dirs

    > Why would you want to sort with series at a higher level than study
    > id? If the data is compliant, there will only every be one entry under
    > each series?


    I _totally_ agree. But unfortunately data that we get is not that
    "compliant".
    Last week got several cases, where several scans were stored into the same
    SeriesID/UID, StudyID/UID, etc, the only way to separate them into coherent
    3D datasets was with AcquisitionTime. Good thing I have explorer shell
    extension
    to sort out the dcm files, but now looking for more semi automatic approach.

    I will have a look at the simple_storage, mentioned in the other reply
    (thanks),
    and I got impatient a bit and wrote arbitrary tag selection sort lib, now
    wrapping
    it into command line interface for scripting.




    Thanks for info.
    Igor

    "Eric Goodall" wrote in message
    news:2701fd0e.0410260837.58bca89a@posting.google.c om...
    > "Igor Okulist" wrote in message

    news:...
    > > Hi,
    > >
    > > What would be "best" tool to sort a number of DICOM files into
    > > PatientID\SeriesID\StudyID\AcquisitionTime directory structure?
    > > (would be nice if I could change/specify the directory hierarchy)
    > >
    > > I could use storescp to resort studies according to UID,
    > > which has the benefit of avoiding possible PatientID, SeriesID clashes
    > > (I have many studies from different places, and would not you know
    > > each is using different scheme for incrementing the IDs, so a clash is
    > > possible), but it does not feel like the right approach.
    > >

    >
    > Why would you want to sort with series at a higher level than study
    > id? If the data is compliant, there will only every be one entry under
    > each series?
    >
    > More to the point, the modality you are using may be generating nice
    > values for Series ID and Study ID, but many do not. DICOM doesn't
    > mandate that either of them be unique. A reasonable interpretation of
    > the semantics of the attributes would be to make Series Ids unique
    > within the scope of the study and study ID unique within some period
    > of time in which multiple studies are generated. Most vendors do a
    > decent job with Study ID, when they bother. Series ID is another
    > story. Very often I've seen a CT or MR study with 6 or 7 series, all
    > labeled with the value "1" for the series ID. You can sort by the IDs,
    > but for some manufactures, all your images are going to end up in the
    > same directory.




  5. Re: sorting dicom files into dirs

    "Igor Okulist" wrote in message news:...
    > Hi,
    >
    > What would be "best" tool to sort a number of DICOM files into
    > PatientID\SeriesID\StudyID\AcquisitionTime directory structure?
    > (would be nice if I could change/specify the directory hierarchy)
    >
    > I could use storescp to resort studies according to UID,
    > which has the benefit of avoiding possible PatientID, SeriesID clashes
    > (I have many studies from different places, and would not you know
    > each is using different scheme for incrementing the IDs, so a clash is
    > possible), but it does not feel like the right approach.
    >
    > Is there such a tool, or should I be writing my own ?
    >
    > Regards,
    > Igor


    Try http://www.geocities.com/gigiobb/UtilityPrograms.txt

  6. Re: sorting dicom files into dirs

    HI,
    Does it make sense of creating a DICOMDIR structure and storing in
    local disk..Is this a standard..But i havent seen this in standard..Is
    this a proper application..

    Regards

    gigiobb@yahoo.com (GG) wrote in message news:...
    > "Igor Okulist" wrote in message news:...
    > > Hi,
    > >
    > > What would be "best" tool to sort a number of DICOM files into
    > > PatientID\SeriesID\StudyID\AcquisitionTime directory structure?
    > > (would be nice if I could change/specify the directory hierarchy)
    > >
    > > I could use storescp to resort studies according to UID,
    > > which has the benefit of avoiding possible PatientID, SeriesID clashes
    > > (I have many studies from different places, and would not you know
    > > each is using different scheme for incrementing the IDs, so a clash is
    > > possible), but it does not feel like the right approach.
    > >
    > > Is there such a tool, or should I be writing my own ?
    > >
    > > Regards,
    > > Igor

    >
    > Try http://www.geocities.com/gigiobb/UtilityPrograms.txt


+ Reply to Thread