DICOM validator updated - DICOM

This is a discussion on DICOM validator updated - DICOM ; Hi all I was updating my DICOM validator (dciodvfy in dicom3tools) to the latest greatest DICOM 2006 edition plus the recent CPs and Supplements, and I took the opportunity to add a number of features, particularly since a large number ...

+ Reply to Thread
Results 1 to 4 of 4

Thread: DICOM validator updated

  1. DICOM validator updated

    Hi all

    I was updating my DICOM validator (dciodvfy in dicom3tools)
    to the latest greatest DICOM 2006 edition plus the recent
    CPs and Supplements, and I took the opportunity to add a
    number of features, particularly since a large number of
    attributes, IODs and modules were retired last year.

    The tool now whines if you use any retired attributes, and
    if you add any extra standard attributes that are not (or
    are no longer) in the IOD. It also checks the number of
    items in sequences (something it should have done a long
    time ago).

    There is also an experimental feature to define "profiles"
    of standard IODs, as IHE has started to do, the only one
    of which defined so far is the IHE Mammography profile;
    this allows you to specify on the command line the (more
    restrictive) profile to apply instead of the standard IOD.

    I would be interested in feedback on these changes, so feel
    free to try it and email me. See:

    http://www.dclunie.com/dicom3tools/workinprogress/

    for the current build.

    David

    PS. Please don't send me "but it doesn't run on Windows"
    feedback ... you might be able to get it going under
    Cygwin if you are lucky, but I haven't tried this in a
    long time.

  2. Re: DICOM validator updated

    David Clunie wrote:
    > Hi all
    >
    > I was updating my DICOM validator (dciodvfy in dicom3tools)
    > to the latest greatest DICOM 2006 edition plus the recent
    > CPs and Supplements, and I took the opportunity to add a
    > number of features, particularly since a large number of
    > attributes, IODs and modules were retired last year.
    >
    > The tool now whines if you use any retired attributes, and
    > if you add any extra standard attributes that are not (or
    > are no longer) in the IOD. It also checks the number of
    > items in sequences (something it should have done a long
    > time ago).
    >
    > There is also an experimental feature to define "profiles"
    > of standard IODs, as IHE has started to do, the only one
    > of which defined so far is the IHE Mammography profile;
    > this allows you to specify on the command line the (more
    > restrictive) profile to apply instead of the standard IOD.
    >
    > I would be interested in feedback on these changes, so feel
    > free to try it and email me. See:
    >
    > http://www.dclunie.com/dicom3tools/workinprogress/
    >
    > for the current build.
    >
    > David
    >
    > PS. Please don't send me "but it doesn't run on Windows"
    > feedback ... you might be able to get it going under
    > Cygwin if you are lucky, but I haven't tried this in a
    > long time.


    David,

    Thanks a bunch for this awsome tool !
    Anyway I think I was not paying attention but when did Group Length
    became retired attributes (*) ? I am looking at section 7.2 (PS 3.5 -
    2006) and I understand they are still Type 3, right ? Doing a google
    search I can find that other people mentioned that it has been retired
    but I cannot find anything from nema.org (**).

    As a side note, redistribution of the toolkit is not authorized but
    solution like for NetBSD:

    http://pkgsrc.se/wip/dicom3tools

    Is fine, right ? I would like -as part of the nightly testing-
    integrate dicom3tools.

    Thanks
    Mathieu
    (*)
    $ ./dciodvfy ./GE_DLX-8-MONO2-Multiframe-j2k.dcm
    Warning - Retired attribute - (0x7fe0,0x0000) UL Pixel Data Group
    Length

    (**)
    Google search:
    http://www.google.com/search?hl=en&l...rg&btnG=Search


  3. Re: DICOM validator updated

    Hi Mathieu

    Mathieu Malaterre wrote:

    > Anyway I think I was not paying attention but when did Group Length
    > became retired attributes (*) ? I am looking at section 7.2 (PS 3.5 -
    > 2006) and I understand they are still Type 3, right ? Doing a google
    > search I can find that other people mentioned that it has been retired
    > but I cannot find anything from nema.org (**).


    Strictly speaking you are correct; they are not formally
    retired, but they should be.

    I will put in a CP.

    In the interim I think I will continue to have the validator
    claim that they are retired, since they are very undesirable.

    If you like, I can change the warning message to say "undesirable"
    instead of "retired"

    > As a side note, redistribution of the toolkit is not authorized but
    > solution like for NetBSD:


    I haven't looked at the copyright notice in a long time, and
    you are right, it is probably excessively restrictive by modern
    standards ... I will change to using the BSD-like license from
    the pixelmed toolkit for the dicom3tools as well.

    David

  4. Re: DICOM validator updated

    Hi David,

    David Clunie wrote:
    > Mathieu Malaterre wrote:
    > Strictly speaking you are correct; they are not formally
    > retired, but they should be.
    >
    > I will put in a CP.


    Could you also put a CP to mark (0029,1010) and (0029,1020) from
    SIEMENS to be undesirable since they contains information encoded in
    binary format that are not available in the public field ;-P

    > In the interim I think I will continue to have the validator
    > claim that they are retired, since they are very undesirable.
    >
    > If you like, I can change the warning message to say "undesirable"
    > instead of "retired"


    I have seen a couple of DICOM where the PixelData length was wrong, so
    we were using the Pixel Data Group Length instead to figure out the
    size of the data. It's sad to see a potential optimization for binary
    file format go away (in XML for instance it is very difficult to use
    this notion of offsets).

    > I haven't looked at the copyright notice in a long time, and
    > you are right, it is probably excessively restrictive by modern
    > standards ... I will change to using the BSD-like license from
    > the pixelmed toolkit for the dicom3tools as well.


    Excellent ! If it becomes BSD style, I should be able to replace the
    current Imake build system with a CMake based one, which -hopefully-
    would make the life of the -poor- Win32 users easier.

    Thanks !
    Mathieu


+ Reply to Thread