problems reading DICOMs - DICOM

This is a discussion on problems reading DICOMs - DICOM ; Hi all, I got a problem reading DICOMs from a CT scans. The images come directly from the CT scans, without the .dcm extension. I assume they are DICOM, since I can read the header with the program dicom2, and ...

+ Reply to Thread
Results 1 to 6 of 6

Thread: problems reading DICOMs

  1. problems reading DICOMs

    Hi all,
    I got a problem reading DICOMs from a CT scans.
    The images come directly from the CT scans, without the .dcm
    extension.
    I assume they are DICOM, since I can read the header with the program
    dicom2, and I can successfully anonymize them with "DICOM Anonymizer
    Light", but I cannot open/see the images with any program I know: i.e.
    I tried ITK-snap, imagej, and also dicom2 is unable to convert the
    files.
    I'm clueless.
    In a first stage I was thinking to a kind of encryption, but actually
    the header is totally clear, with even the name of the patient
    visible. Raw data? I don't think so: IMG Window Center and Width are
    setted, also the values of "IMG Rows" and "IMG columns" are setted.
    In the other hands, from the proprietary software (Philips) it is
    possible to see the images.
    I'm driving crazy, anyone else encountered a similar problem?

    cheers,
    Mauro

  2. Re: problems reading DICOMs

    On Apr 23, 11:00 am, mauro wrote:
    > Hi all,
    > I got a problem reading DICOMs from a CT scans.
    > The images come directly from the CT scans, without the .dcm
    > extension.
    > I assume they are DICOM, since I can read the header with the program
    > dicom2, and I can successfully anonymize them with "DICOM Anonymizer
    > Light", but I cannot open/see the images with any program I know: i.e.
    > I tried ITK-snap, imagej, and also dicom2 is unable to convert the
    > files.
    > I'm clueless.
    > In a first stage I was thinking to a kind of encryption, but actually
    > the header is totally clear, with even the name of the patient
    > visible. Raw data? I don't think so: IMG Window Center and Width are
    > setted, also the values of "IMG Rows" and "IMG columns" are setted.
    > In the other hands, from the proprietary software (Philips) it is
    > possible to see the images.
    > I'm driving crazy, anyone else encountered a similar problem?


    Hi Mauro,

    You are dealing with the well-known ELSCINT LOSSLESS compressed
    image. One way to recognize them (eventhough they are incorrectly
    reported as 1.2.840.10008.5.1.4.1.1.2 [CT Image Storage]), is to
    check for the presence of 7fe0,0010: the Pixel Data element. It is
    missing from your image, but instead you can find a data element
    (07a1,100a) which AFAIK is the place holder for the compressed Pixel
    Data bytes.

    I was about to recommend you to either:
    1. Ask your PACS admin to correctly export the image via the console,
    so that you get a correct (conforming) DICOM file
    2. If you are stuck and have no access to the console, you have some
    intermediate solution like TomoVision

    HTH
    -Mathieu
    Ps: it is actually a pain to recognize them, but I'll try to fix ITK-
    snap which is underneath calling GDCM so that at least we report a
    meaningful error message.

  3. Re: problems reading DICOMs

    Hi Mathieu,
    first of all, thanks for the useful information!

    > 2. If you are stuck and have no access to the console, you have some
    > intermediate solution like TomoVision


    this is exactly my case. TomoVision is an excellent program, it
    recognize the ELSCINT LOSSLESS, but unfortunately it actually doesn't
    recover my images.
    Do you have any clue about other solutions? I also unsuccessfully
    tried xmedcon...

    > Ps: it is actually a pain to recognize them, but I'll try to fix ITK-
    > snap which is underneath calling GDCM so that at least we report a
    > meaningful error message.


    indeed ;-)

    cheers,
    Mauro


  4. Re: problems reading DICOMs

    On Apr 23, 3:27 pm, mauro wrote:
    > Hi Mathieu,
    > first of all, thanks for the useful information!
    >
    > > 2. If you are stuck and have no access to the console, you have some
    > > intermediate solution like TomoVision

    >
    > this is exactly my case. TomoVision is an excellent program, it
    > recognize the ELSCINT LOSSLESS, but unfortunately it actually doesn't
    > recover my images.
    > Do you have any clue about other solutions? I also unsuccessfully
    > tried xmedcon...


    xmedcon is open source and as such does not have access to the
    proprietary information required to decompressed the binary blob at
    element with tag (07a1,0a,ELSINT1)
    You need to wait for Yves to give a complete answer, but it looks like
    that this is not a true 'LOSSLESS RICE' image as I have previously
    seen, look carefully at the header and you'll see that element with
    tag (07a1,11,ELSCINT1) is in fact PMSCT_RLE1 (which I have never seen
    before).

    ....
    (07a1,0010) LO [ELSCINT1] #
    8,1 Private Creator
    (07a1,100a) OB 5a\92\00\ef\9b\e4\00\00... # 1316426,1 Tamar
    Compressed Pixel Data
    (07a1,1010) LO [3.5 ] #
    4,1 Tamar Software Version
    (07a1,1011) CS [PMSCT_RLE1] #
    10,1 Tamar Compression Type
    (7fdf,0010) LO [ELSCINT1] #
    8,1 Private Creator
    (7fdf,10f0) OB (no value) #
    0,1 ?
    (7fdf,10ff) SH (no value) #
    0,1 ?

    HTH
    -Mathieu

  5. Re: problems reading DICOMs

    Bonjour Mauro,

    I just added your Philips compression format to my programs. You can
    now use TomoVision to display the images and DICOMatic to convert them
    to normal (uncomressed) DICOM.

    Merci,
    Yves

    >On Apr 23, 3:27 pm, mauro wrote:
    >> Hi Mathieu,
    >> first of all, thanks for the useful information!
    >>
    >> > 2. If you are stuck and have no access to the console, you have some
    >> > intermediate solution like TomoVision

    >>
    >> this is exactly my case. TomoVision is an excellent program, it
    >> recognize the ELSCINT LOSSLESS, but unfortunately it actually doesn't
    >> recover my images.
    >> Do you have any clue about other solutions? I also unsuccessfully
    >> tried xmedcon...

    >
    >xmedcon is open source and as such does not have access to the
    >proprietary information required to decompressed the binary blob at
    >element with tag (07a1,0a,ELSINT1)
    >You need to wait for Yves to give a complete answer, but it looks like
    >that this is not a true 'LOSSLESS RICE' image as I have previously
    >seen, look carefully at the header and you'll see that element with
    >tag (07a1,11,ELSCINT1) is in fact PMSCT_RLE1 (which I have never seen
    >before).
    >
    >...
    >(07a1,0010) LO [ELSCINT1] #
    >8,1 Private Creator
    >(07a1,100a) OB 5a\92\00\ef\9b\e4\00\00... # 1316426,1 Tamar
    >Compressed Pixel Data
    >(07a1,1010) LO [3.5 ] #
    >4,1 Tamar Software Version
    >(07a1,1011) CS [PMSCT_RLE1] #
    >10,1 Tamar Compression Type
    >(7fdf,0010) LO [ELSCINT1] #
    >8,1 Private Creator
    >(7fdf,10f0) OB (no value) #
    >0,1 ?
    >(7fdf,10ff) SH (no value) #
    >0,1 ?
    >
    >HTH
    >-Mathieu


  6. Re: problems reading DICOMs

    Mauro,

    GDCM 2.x now also supports this image. this will be part of the GDCM
    2.0.5 release.

    -Mathieu

    On Apr 28, 11:59 pm, Yves Martel wrote:
    > Bonjour Mauro,
    >
    > I just added your Philips compression format to my programs. You can
    > now use TomoVision to display the images and DICOMatic to convert them
    > to normal (uncomressed) DICOM.
    >
    > Merci,
    > Yves
    >
    > >On Apr 23, 3:27 pm, mauro wrote:
    > >> Hi Mathieu,
    > >> first of all, thanks for the useful information!

    >
    > >> > 2. If you are stuck and have no access to the console, you have some
    > >> > intermediate solution like TomoVision

    >
    > >> this is exactly my case. TomoVision is an excellent program, it
    > >> recognize the ELSCINT LOSSLESS, but unfortunately it actually doesn't
    > >> recover my images.
    > >> Do you have any clue about other solutions? I also unsuccessfully
    > >> tried xmedcon...

    >
    > >xmedcon is open source and as such does not have access to the
    > >proprietary information required to decompressed the binary blob at
    > >element with tag (07a1,0a,ELSINT1)
    > >You need to wait for Yves to give a complete answer, but it looks like
    > >that this is not a true 'LOSSLESS RICE' image as I have previously
    > >seen, look carefully at the header and you'll see that element with
    > >tag (07a1,11,ELSCINT1) is in fact PMSCT_RLE1 (which I have never seen
    > >before).

    >
    > >...
    > >(07a1,0010) LO [ELSCINT1] #
    > >8,1 Private Creator
    > >(07a1,100a) OB 5a\92\00\ef\9b\e4\00\00... # 1316426,1 Tamar
    > >Compressed Pixel Data
    > >(07a1,1010) LO [3.5 ] #
    > >4,1 Tamar Software Version
    > >(07a1,1011) CS [PMSCT_RLE1] #
    > >10,1 Tamar Compression Type
    > >(7fdf,0010) LO [ELSCINT1] #
    > >8,1 Private Creator
    > >(7fdf,10f0) OB (no value) #
    > >0,1 ?
    > >(7fdf,10ff) SH (no value) #
    > >0,1 ?

    >
    > >HTH
    > >-Mathieu



+ Reply to Thread