Separate DICOM overlays and SOP Instance UID - DICOM

This is a discussion on Separate DICOM overlays and SOP Instance UID - DICOM ; Hi: I have a question about when you need to create a SOP Instance UID for an image. I looked at the DICOM standard 2003, part 3 - C.7.6.1.1.3. Examples of Derived Images that would normally be expected to affect ...

+ Reply to Thread
Results 1 to 4 of 4

Thread: Separate DICOM overlays and SOP Instance UID

  1. Separate DICOM overlays and SOP Instance UID

    Hi:

    I have a question about when you need to create a SOP Instance UID for
    an image.

    I looked at the DICOM standard 2003, part 3 - C.7.6.1.1.3.

    Examples of Derived Images that would normally be expected to affect
    professional
    interpretation and would thus have a new UID include:
    a. images resulting from image processing of another image (e.g.
    unsharp masking),
    b. a multiplanar reformatted CT image,
    c. a DSA image derived by subtracting pixel values of one image from
    another.
    d. an image that has been decompressed after having been compressed
    with a lossy
    compression algorithm. To ensure that the user has the necessary
    information about the
    lossy compression, the approximate compression ratio may be included in
    Derivation
    Description (0008,2111).

    But it is not clear to me if a new SOP Instance UID should be created
    for the following workflow:

    1. PACS sends image to workstation
    2. Workstation adds annotations/graphics to image and converts the
    annotation to separate DICOM overlay. e.g the graphic data are sent in
    Overlay Data Element (60xx,3000) so it did not change the Pixel Data
    Element (7FE0,0010).

    Since PACS may choose to ignore the image if the SOP Instance UID
    already exist in PACS. As a result, the clinical user will not see the
    annotations next time he/she views the image from PACS.

    Thanks

    Wayne Tran


  2. Re: Separate DICOM overlays and SOP Instance UID

    Hi:

    I forgot to give step (3).

    1. PACS sends image to workstation
    2. Workstation adds annotations/graphics to image and converts the
    annotation to separate DICOM overlay. e.g the graphic data are sent in
    Overlay Data Element (60xx,3000) so it did not change the Pixel Data
    Element (7FE0,0010).

    3. The workstation sends the image with separate DICOM overlay back to
    PACS.

    Should the workstation create a new SOP Instance UID for it?

    Since PACS may choose to ignore the image if the SOP Instance UID
    already exist in PACS. As a result, the clinical user will not see the
    annotations next time he/she views the image from PACS.


    Thanks

    Wayne Tran


    enyaw_2010@hotmail.com wrote:
    > Hi:
    >
    > I have a question about when you need to create a SOP Instance UID for
    > an image.
    >
    > I looked at the DICOM standard 2003, part 3 - C.7.6.1.1.3.
    >
    > Examples of Derived Images that would normally be expected to affect
    > professional
    > interpretation and would thus have a new UID include:
    > a. images resulting from image processing of another image (e.g.
    > unsharp masking),
    > b. a multiplanar reformatted CT image,
    > c. a DSA image derived by subtracting pixel values of one image from
    > another.
    > d. an image that has been decompressed after having been compressed
    > with a lossy
    > compression algorithm. To ensure that the user has the necessary
    > information about the
    > lossy compression, the approximate compression ratio may be included in
    > Derivation
    > Description (0008,2111).
    >
    > But it is not clear to me if a new SOP Instance UID should be created
    > for the following workflow:
    >
    > 1. PACS sends image to workstation
    > 2. Workstation adds annotations/graphics to image and converts the
    > annotation to separate DICOM overlay. e.g the graphic data are sent in
    > Overlay Data Element (60xx,3000) so it did not change the Pixel Data
    > Element (7FE0,0010).
    >
    > Since PACS may choose to ignore the image if the SOP Instance UID
    > already exist in PACS. As a result, the clinical user will not see the
    > annotations next time he/she views the image from PACS.
    >
    > Thanks
    >
    > Wayne Tran



  3. Re: Separate DICOM overlays and SOP Instance UID

    Hi Wayne

    The best thing would be NOT TO DO THIS AT ALL.

    It is very bad practice to modify images by adding overlays.

    That is what presentation states are for.

    But if you do it, there is absolutely no question that you
    should assign a new SOP Instance UID - if the overlay contains
    information worth preserving then it almost certainly would
    'affect professional interpretation' (stupid rule, I know,
    it should be less ambiguous than this, but that is history).

    Also, pragmatically, as you point out, new UIDs are the only way
    to be sure that the PACS doesn't discard it, though there will
    then be two "versions" with the same pixel data; further, you will
    need to put the new images in a new Series (based on the different
    equipment rule), which at least makes it easy to distinguish them.

    Finally, many devices will not render overlays in images (whether
    they be embedded in (7FE0,0010) or in (60xx,3000)), so presentation
    states are much more reliable for this (in the sense that a PACS
    either does or does not support them, unambiguously).

    All good reasons not to do what you are describing.

    In short ... overlays in images suck ... do not use them.

    David

    enyaw_2010@hotmail.com wrote:
    > Hi:
    >
    > I forgot to give step (3).
    >
    > 1. PACS sends image to workstation
    > 2. Workstation adds annotations/graphics to image and converts the
    > annotation to separate DICOM overlay. e.g the graphic data are sent in
    > Overlay Data Element (60xx,3000) so it did not change the Pixel Data
    > Element (7FE0,0010).
    >
    > 3. The workstation sends the image with separate DICOM overlay back to
    > PACS.
    >
    > Should the workstation create a new SOP Instance UID for it?
    >
    > Since PACS may choose to ignore the image if the SOP Instance UID
    > already exist in PACS. As a result, the clinical user will not see the
    > annotations next time he/she views the image from PACS.
    >
    >
    > Thanks
    >
    > Wayne Tran
    >
    >
    > enyaw_2010@hotmail.com wrote:
    >> Hi:
    >>
    >> I have a question about when you need to create a SOP Instance UID for
    >> an image.
    >>
    >> I looked at the DICOM standard 2003, part 3 - C.7.6.1.1.3.
    >>
    >> Examples of Derived Images that would normally be expected to affect
    >> professional
    >> interpretation and would thus have a new UID include:
    >> a. images resulting from image processing of another image (e.g.
    >> unsharp masking),
    >> b. a multiplanar reformatted CT image,
    >> c. a DSA image derived by subtracting pixel values of one image from
    >> another.
    >> d. an image that has been decompressed after having been compressed
    >> with a lossy
    >> compression algorithm. To ensure that the user has the necessary
    >> information about the
    >> lossy compression, the approximate compression ratio may be included in
    >> Derivation
    >> Description (0008,2111).
    >>
    >> But it is not clear to me if a new SOP Instance UID should be created
    >> for the following workflow:
    >>
    >> 1. PACS sends image to workstation
    >> 2. Workstation adds annotations/graphics to image and converts the
    >> annotation to separate DICOM overlay. e.g the graphic data are sent in
    >> Overlay Data Element (60xx,3000) so it did not change the Pixel Data
    >> Element (7FE0,0010).
    >>
    >> Since PACS may choose to ignore the image if the SOP Instance UID
    >> already exist in PACS. As a result, the clinical user will not see the
    >> annotations next time he/she views the image from PACS.
    >>
    >> Thanks
    >>
    >> Wayne Tran

    >


  4. Re: Separate DICOM overlays and SOP Instance UID

    Hi Dave:

    Thanks for the clarification/explanation.

    Wayne Tran

    David Clunie wrote:
    > Hi Wayne
    >
    > The best thing would be NOT TO DO THIS AT ALL.
    >
    > It is very bad practice to modify images by adding overlays.
    >
    > That is what presentation states are for.
    >
    > But if you do it, there is absolutely no question that you
    > should assign a new SOP Instance UID - if the overlay contains
    > information worth preserving then it almost certainly would
    > 'affect professional interpretation' (stupid rule, I know,
    > it should be less ambiguous than this, but that is history).
    >
    > Also, pragmatically, as you point out, new UIDs are the only way
    > to be sure that the PACS doesn't discard it, though there will
    > then be two "versions" with the same pixel data; further, you will
    > need to put the new images in a new Series (based on the different
    > equipment rule), which at least makes it easy to distinguish them.
    >
    > Finally, many devices will not render overlays in images (whether
    > they be embedded in (7FE0,0010) or in (60xx,3000)), so presentation
    > states are much more reliable for this (in the sense that a PACS
    > either does or does not support them, unambiguously).
    >
    > All good reasons not to do what you are describing.
    >
    > In short ... overlays in images suck ... do not use them.
    >
    > David
    >
    > enyaw_2010@hotmail.com wrote:
    > > Hi:
    > >
    > > I forgot to give step (3).
    > >
    > > 1. PACS sends image to workstation
    > > 2. Workstation adds annotations/graphics to image and converts the
    > > annotation to separate DICOM overlay. e.g the graphic data are sent in
    > > Overlay Data Element (60xx,3000) so it did not change the Pixel Data
    > > Element (7FE0,0010).
    > >
    > > 3. The workstation sends the image with separate DICOM overlay back to
    > > PACS.
    > >
    > > Should the workstation create a new SOP Instance UID for it?
    > >
    > > Since PACS may choose to ignore the image if the SOP Instance UID
    > > already exist in PACS. As a result, the clinical user will not see the
    > > annotations next time he/she views the image from PACS.
    > >
    > >
    > > Thanks
    > >
    > > Wayne Tran
    > >
    > >
    > > enyaw_2010@hotmail.com wrote:
    > >> Hi:
    > >>
    > >> I have a question about when you need to create a SOP Instance UID for
    > >> an image.
    > >>
    > >> I looked at the DICOM standard 2003, part 3 - C.7.6.1.1.3.
    > >>
    > >> Examples of Derived Images that would normally be expected to affect
    > >> professional
    > >> interpretation and would thus have a new UID include:
    > >> a. images resulting from image processing of another image (e.g.
    > >> unsharp masking),
    > >> b. a multiplanar reformatted CT image,
    > >> c. a DSA image derived by subtracting pixel values of one image from
    > >> another.
    > >> d. an image that has been decompressed after having been compressed
    > >> with a lossy
    > >> compression algorithm. To ensure that the user has the necessary
    > >> information about the
    > >> lossy compression, the approximate compression ratio may be included in
    > >> Derivation
    > >> Description (0008,2111).
    > >>
    > >> But it is not clear to me if a new SOP Instance UID should be created
    > >> for the following workflow:
    > >>
    > >> 1. PACS sends image to workstation
    > >> 2. Workstation adds annotations/graphics to image and converts the
    > >> annotation to separate DICOM overlay. e.g the graphic data are sent in
    > >> Overlay Data Element (60xx,3000) so it did not change the Pixel Data
    > >> Element (7FE0,0010).
    > >>
    > >> Since PACS may choose to ignore the image if the SOP Instance UID
    > >> already exist in PACS. As a result, the clinical user will not see the
    > >> annotations next time he/she views the image from PACS.
    > >>
    > >> Thanks
    > >>
    > >> Wayne Tran

    > >



+ Reply to Thread