Study ID absent in DICOMDIR - DICOM

This is a discussion on Study ID absent in DICOMDIR - DICOM ; I have a DICOM viewer application which can read studies from US machines & has an CD archive option. I recently figured out that Study ID (0020 , 0010) is a Type 1 attribute in case of DICOMDIR. But there ...

+ Reply to Thread
Results 1 to 4 of 4

Thread: Study ID absent in DICOMDIR

  1. Study ID absent in DICOMDIR

    I have a DICOM viewer application which can read studies from US
    machines & has an CD archive option. I recently figured out that Study
    ID (0020 , 0010) is a Type 1 attribute in case of DICOMDIR.

    But there are some studies (eg. GE logic images) where the Study ID has
    zero length in the DICOMDIR inspite of being Type 1. If the user only
    wishes to load the the study in my application, delete some of the
    images & re-create the study using my archival module, I modify the
    Implementation version name in the DICOMDIR & archive it. But I have a
    check for the TYPE 1 attributes & hence archival of this study fails.

    1. If I allow the user to proceed with the archival even when study ID
    (or for that matter any Type 1 field) is missing, am I violating the
    standard?

    2. I could alternatively generate my own study ID & change the DICOMDIR
    but there would be a mismatch with the study ID tag in the images that
    are being archived.

    Please suggest what I should do in this case.

    What could be the reason that the same tag is a Type 2 attribute for
    the general study module in an image but a Type 1 in case of dicomdir?


  2. Re: Study ID absent in DICOMDIR

    Hi Supraja

    I will try to answer your query as per my understanding one by one :

    I have a DICOM viewer application which can read studies from US
    machines & has an CD archive option. I recently figured out that Study
    ID (0020 , 0010) is a Type 1 attribute in case of DICOMDIR.

    But there are some studies (eg. GE logic images) where the Study ID has
    zero length in the DICOMDIR inspite of being Type 1. If the user only
    wishes to load the the study in my application, delete some of the
    images & re-create the study using my archival module, I modify the
    Implementation version name in the DICOMDIR & archive it. But I have a
    check for the TYPE 1 attributes & hence archival of this study fails.

    1. If I allow the user to proceed with the archival even when study ID
    (or for that matter any Type 1 field) is missing, am I violating the
    standard?

    -----> Yes it should be the violation of protocol.

    2. I could alternatively generate my own study ID & change the DICOMDIR
    but there would be a mismatch with the study ID tag in the images that
    are being archived. Please suggest what I should do in this case.

    -----> Study ID in the record has to be same as that in the images you
    are archiving. The images belonging to the same study will have same
    Study ID as that in the study record. You can pick up Study ID from the
    image(s) belonging to the same study and place it in the directory
    record. If its value is missing in the image then I guess you should
    generate a study ID update the dataset and directory record.

    What could be the reason that the same tag is a Type 2 attribute for
    the general study module in an image but a Type 1 in case of dicomdir?

    -----> If we talk about General study module then Study ID is one
    attribute of that module.

    In case if DICOMDIR we talk about directory records, there are keys for
    those records and depending on the application profiles some of the
    Type 2 and Type 3 can be moved to Type 1. In addition to this you can
    have your private keys also.

    In DICOMDIR, study instance UID is Type 1C whereas in General study
    module it is Type 1. Study instance UID may or may not be there in
    DICOMDIR.

    >From part 3 "The Study Instance UID shall be present as a mandatory key

    only if no file is referenced by this Directory Record. In the case
    where this Directory Record references a file, the Directory Record
    contains in the Referenced SOP Instance UID in File (0004,1511). In
    this case (0004,1511) may be used as a "pseudo" Directory Record Key
    (See Table F.3-3) and need not be duplicated."

    Please refer to the answer of question 2. I guess it is the reason why
    Study ID is type 1.

    I hope this helps. If anyone feels that my understanding is wrong then
    please correct me.

    Regards
    Bimalendu



    supraja wrote:
    > I have a DICOM viewer application which can read studies from US
    > machines & has an CD archive option. I recently figured out that Study
    > ID (0020 , 0010) is a Type 1 attribute in case of DICOMDIR.
    >
    > But there are some studies (eg. GE logic images) where the Study ID has
    > zero length in the DICOMDIR inspite of being Type 1. If the user only
    > wishes to load the the study in my application, delete some of the
    > images & re-create the study using my archival module, I modify the
    > Implementation version name in the DICOMDIR & archive it. But I have a
    > check for the TYPE 1 attributes & hence archival of this study fails.
    >
    > 1. If I allow the user to proceed with the archival even when study ID
    > (or for that matter any Type 1 field) is missing, am I violating the
    > standard?
    >
    > 2. I could alternatively generate my own study ID & change the DICOMDIR
    > but there would be a mismatch with the study ID tag in the images that
    > are being archived.
    >
    > Please suggest what I should do in this case.
    >
    > What could be the reason that the same tag is a Type 2 attribute for
    > the general study module in an image but a Type 1 in case of dicomdir?



  3. Re: Study ID absent in DICOMDIR

    Hi Bimalendu,

    Thanks for the explanation.

    > 1. If I allow the user to proceed with the archival even when study ID
    > (or for that matter any Type 1 field) is missing, am I violating the
    > standard?
    >
    > -----> Yes it should be the violation of protocol.


    I am proceeding with the CD write but I display a message now stating
    that the tag value is invalid. It would be quite an effort to create my
    own study id each time & update the dicomdir & all the images that it
    references. Besides, study id is not the only such tag. There are other
    tags which might be missing & I'll end up doing lot of work.

    Thanks,
    Supraja

    bimalendug@gmail.com wrote:
    > Hi Supraja
    >
    > I will try to answer your query as per my understanding one by one :
    >
    > I have a DICOM viewer application which can read studies from US
    > machines & has an CD archive option. I recently figured out that Study
    > ID (0020 , 0010) is a Type 1 attribute in case of DICOMDIR.
    >
    > But there are some studies (eg. GE logic images) where the Study ID has
    > zero length in the DICOMDIR inspite of being Type 1. If the user only
    > wishes to load the the study in my application, delete some of the
    > images & re-create the study using my archival module, I modify the
    > Implementation version name in the DICOMDIR & archive it. But I have a
    > check for the TYPE 1 attributes & hence archival of this study fails.
    >
    > 1. If I allow the user to proceed with the archival even when study ID
    > (or for that matter any Type 1 field) is missing, am I violating the
    > standard?
    >
    > -----> Yes it should be the violation of protocol.
    >
    > 2. I could alternatively generate my own study ID & change the DICOMDIR
    > but there would be a mismatch with the study ID tag in the images that
    > are being archived. Please suggest what I should do in this case.
    >
    > -----> Study ID in the record has to be same as that in the images you
    > are archiving. The images belonging to the same study will have same
    > Study ID as that in the study record. You can pick up Study ID from the
    > image(s) belonging to the same study and place it in the directory
    > record. If its value is missing in the image then I guess you should
    > generate a study ID update the dataset and directory record.
    >
    > What could be the reason that the same tag is a Type 2 attribute for
    > the general study module in an image but a Type 1 in case of dicomdir?
    >
    > -----> If we talk about General study module then Study ID is one
    > attribute of that module.
    >
    > In case if DICOMDIR we talk about directory records, there are keys for
    > those records and depending on the application profiles some of the
    > Type 2 and Type 3 can be moved to Type 1. In addition to this you can
    > have your private keys also.
    >
    > In DICOMDIR, study instance UID is Type 1C whereas in General study
    > module it is Type 1. Study instance UID may or may not be there in
    > DICOMDIR.
    >
    > >From part 3 "The Study Instance UID shall be present as a mandatory key

    > only if no file is referenced by this Directory Record. In the case
    > where this Directory Record references a file, the Directory Record
    > contains in the Referenced SOP Instance UID in File (0004,1511). In
    > this case (0004,1511) may be used as a "pseudo" Directory Record Key
    > (See Table F.3-3) and need not be duplicated."
    >
    > Please refer to the answer of question 2. I guess it is the reason why
    > Study ID is type 1.
    >
    > I hope this helps. If anyone feels that my understanding is wrong then
    > please correct me.
    >
    > Regards
    > Bimalendu
    >
    >
    >
    > supraja wrote:
    > > I have a DICOM viewer application which can read studies from US
    > > machines & has an CD archive option. I recently figured out that Study
    > > ID (0020 , 0010) is a Type 1 attribute in case of DICOMDIR.
    > >
    > > But there are some studies (eg. GE logic images) where the Study ID has
    > > zero length in the DICOMDIR inspite of being Type 1. If the user only
    > > wishes to load the the study in my application, delete some of the
    > > images & re-create the study using my archival module, I modify the
    > > Implementation version name in the DICOMDIR & archive it. But I have a
    > > check for the TYPE 1 attributes & hence archival of this study fails.
    > >
    > > 1. If I allow the user to proceed with the archival even when study ID
    > > (or for that matter any Type 1 field) is missing, am I violating the
    > > standard?
    > >
    > > 2. I could alternatively generate my own study ID & change the DICOMDIR
    > > but there would be a mismatch with the study ID tag in the images that
    > > are being archived.
    > >
    > > Please suggest what I should do in this case.
    > >
    > > What could be the reason that the same tag is a Type 2 attribute for
    > > the general study module in an image but a Type 1 in case of dicomdir?



  4. Re: Study ID absent in DICOMDIR

    Hi Supraja

    "Study ID could be user or equipment generated" so I guess you can
    handle the situation where you might need to repair the DICOMDIR which
    is missing Study ID, again its upto you how you want to handle the
    situation, you may raise error also.

    It might not be true for other attributes of directory records. I guess
    there you will have to raise error for missing/invalid attribute.

    Regards
    Bimalendu

    supraja wrote:
    > Hi Bimalendu,
    >
    > Thanks for the explanation.
    >
    > > 1. If I allow the user to proceed with the archival even when study ID
    > > (or for that matter any Type 1 field) is missing, am I violating the
    > > standard?
    > >
    > > -----> Yes it should be the violation of protocol.

    >
    > I am proceeding with the CD write but I display a message now stating
    > that the tag value is invalid. It would be quite an effort to create my
    > own study id each time & update the dicomdir & all the images that it
    > references. Besides, study id is not the only such tag. There are other
    > tags which might be missing & I'll end up doing lot of work.
    >
    > Thanks,
    > Supraja
    >
    > bimalendug@gmail.com wrote:
    > > Hi Supraja
    > >
    > > I will try to answer your query as per my understanding one by one :
    > >
    > > I have a DICOM viewer application which can read studies from US
    > > machines & has an CD archive option. I recently figured out that Study
    > > ID (0020 , 0010) is a Type 1 attribute in case of DICOMDIR.
    > >
    > > But there are some studies (eg. GE logic images) where the Study ID has
    > > zero length in the DICOMDIR inspite of being Type 1. If the user only
    > > wishes to load the the study in my application, delete some of the
    > > images & re-create the study using my archival module, I modify the
    > > Implementation version name in the DICOMDIR & archive it. But I have a
    > > check for the TYPE 1 attributes & hence archival of this study fails.
    > >
    > > 1. If I allow the user to proceed with the archival even when study ID
    > > (or for that matter any Type 1 field) is missing, am I violating the
    > > standard?
    > >
    > > -----> Yes it should be the violation of protocol.
    > >
    > > 2. I could alternatively generate my own study ID & change the DICOMDIR
    > > but there would be a mismatch with the study ID tag in the images that
    > > are being archived. Please suggest what I should do in this case.
    > >
    > > -----> Study ID in the record has to be same as that in the images you
    > > are archiving. The images belonging to the same study will have same
    > > Study ID as that in the study record. You can pick up Study ID from the
    > > image(s) belonging to the same study and place it in the directory
    > > record. If its value is missing in the image then I guess you should
    > > generate a study ID update the dataset and directory record.
    > >
    > > What could be the reason that the same tag is a Type 2 attribute for
    > > the general study module in an image but a Type 1 in case of dicomdir?
    > >
    > > -----> If we talk about General study module then Study ID is one
    > > attribute of that module.
    > >
    > > In case if DICOMDIR we talk about directory records, there are keys for
    > > those records and depending on the application profiles some of the
    > > Type 2 and Type 3 can be moved to Type 1. In addition to this you can
    > > have your private keys also.
    > >
    > > In DICOMDIR, study instance UID is Type 1C whereas in General study
    > > module it is Type 1. Study instance UID may or may not be there in
    > > DICOMDIR.
    > >
    > > >From part 3 "The Study Instance UID shall be present as a mandatory key

    > > only if no file is referenced by this Directory Record. In the case
    > > where this Directory Record references a file, the Directory Record
    > > contains in the Referenced SOP Instance UID in File (0004,1511). In
    > > this case (0004,1511) may be used as a "pseudo" Directory Record Key
    > > (See Table F.3-3) and need not be duplicated."
    > >
    > > Please refer to the answer of question 2. I guess it is the reason why
    > > Study ID is type 1.
    > >
    > > I hope this helps. If anyone feels that my understanding is wrong then
    > > please correct me.
    > >
    > > Regards
    > > Bimalendu
    > >
    > >
    > >
    > > supraja wrote:
    > > > I have a DICOM viewer application which can read studies from US
    > > > machines & has an CD archive option. I recently figured out that Study
    > > > ID (0020 , 0010) is a Type 1 attribute in case of DICOMDIR.
    > > >
    > > > But there are some studies (eg. GE logic images) where the Study ID has
    > > > zero length in the DICOMDIR inspite of being Type 1. If the user only
    > > > wishes to load the the study in my application, delete some of the
    > > > images & re-create the study using my archival module, I modify the
    > > > Implementation version name in the DICOMDIR & archive it. But I have a
    > > > check for the TYPE 1 attributes & hence archival of this study fails.
    > > >
    > > > 1. If I allow the user to proceed with the archival even when study ID
    > > > (or for that matter any Type 1 field) is missing, am I violating the
    > > > standard?
    > > >
    > > > 2. I could alternatively generate my own study ID & change the DICOMDIR
    > > > but there would be a mismatch with the study ID tag in the images that
    > > > are being archived.
    > > >
    > > > Please suggest what I should do in this case.
    > > >
    > > > What could be the reason that the same tag is a Type 2 attribute for
    > > > the general study module in an image but a Type 1 in case of dicomdir?



+ Reply to Thread