Fluoroscopy images and display shutters - DICOM

This is a discussion on Fluoroscopy images and display shutters - DICOM ; Hi all Today's crusade is about Display Shutters. I am doing a little survey as a consequence of an incompatibility between two systems (a fluoroscopy modality from one vendor and a PACS from another vendor) reported by a site (see ...

+ Reply to Thread
Results 1 to 2 of 2

Thread: Fluoroscopy images and display shutters

  1. Fluoroscopy images and display shutters

    Hi all

    Today's crusade is about Display Shutters.

    I am doing a little survey as a consequence of an incompatibility
    between two systems (a fluoroscopy modality from one vendor and
    a PACS from another vendor) reported by a site (see Vince's
    earlier post to pacs_admin for details if you are interested).

    How many PACS support application of Display Shutters encoded
    within images on their workstations or web clients ?

    I took one of the site's RF images that contains
    display shutters and has "white" pixels underneath where the
    X-ray beam has been collimated off, and put it on my web site
    for anyone who is interested to download and try:

    http://www.dclunie.com/images/Fluoro...hutter.dcm.zip

    and let me know what is the result of displaying this normally,
    and inverting it, together with their PACS make, model and
    version.

    This is a de-identified image but from a real product, so it does
    not test all permutations and combinations of shutters as an
    exhaustive test set might do, but it does serve to illustrate the
    practical problem described further below.

    The DICOM Standard defines a Display Shutter, a means of
    indicating to the display where the background should be
    suppressed, regardless of whether or not the image is displayed
    normally or inverted (i.e., bone and contrast are white or black).
    It may be circular, rectangular or polygonal or a combination
    of all three.

    If this is the means chosen by the modality to suppress the
    background, and the pixels underneath are "white" in the
    normal display mode (e.g., if they were collimated off and
    the detector not exposed), then if the display does not support
    the shutter then this will result in a mass of white perimeter
    blinding the user.

    "Burning in" the shutter as a particular value at the modality
    (e.g., "black" instead ) doesn't really help much, since if one
    inverts a white bone or contrast image to get a black bone or
    contrast image then the background will again go to white.

    The need to support this feature has been discussed off and
    on for over a decade of course, as workstations and PACS
    have had to support fluoroscopy and angiography modalities,
    though this is not to say that shutters are not also useful
    for plain projection radiography (CR and digital x-ray) as
    well.

    For example, this issue was discussed during the development
    of the VA/DOD Modality Requirements; see 4.4.4 Pixel Representation
    in:

    http://www1.va.gov/imaging/docs/Join..._V_3_0_upd.pdf

    It specifically mentions that "display workstations
    should not invert the area outside the shuttered area when
    the image is inverted", though the document does not actually
    specify PACS requirements. The initial work on that document
    probably predates IHE, so it is not as if it is something
    new that should not have been supported a long time ago.

    One would have thought that every contemporary PACS would
    have dealt with this, but I am told that currently deployed
    versions of some do not (though recent upgrades, at non-trivial
    cost of course, apparently rectify this problem).

    Before eviscerating any particular vendor for being
    particularly lame or belated in this regard, I wanted to find
    out just how many other systems have this problem, or indeed
    if users actually perceive it to be a significant problem.

    David

    PS. Display shutters are also specified in Presentation States,
    and they are required to be supported by the standard - in
    theory any PACS that supports GSPS should support Display
    Shutters but I am wondering how many actually do in products;
    the IHE MESA tests for Image Display functionality specifically
    test for this for actors in the CPI profile. Even if a PACS
    supports this actor and profile though, and applies display
    shutters in presentation states, that does not necessarily
    mean that they will apply those found encoded in images.


  2. Fluoroscopy images and display shutters - survey results

    Hi all

    Thanks everyone who participated in this survey.

    The results are summarized below; not exactly a comprehensive
    coverage of all systems out there, but clearly some folks have
    got this right, others have only recently got this right and
    need you to upgrade, and others still haven't got this right.

    I will be happy to update or correct these results should any
    further information come to hand.

    GE Centricity:

    - 1.0 CSR5 - BAD - white background
    - 2.1.2.1 - BAD - white background
    - 3.0 - GOOD - black background (default and inverted)

    Siemens PACS

    - V50 - GOOD - black background (default and inverted)

    Fuji synapse

    - v3.0 - BAD - white background
    - v3.1 - BAD - white background
    - v3.11 - BAD - white background

    AGFA Impax

    - 6.0 EE (AGFA/TIANI JVISION 3.3.16) - GOOD - black background (default and inverted)

    SmartPACS

    - 5.0.8.1 - GOOD - black background (default and inverted)

    Philips/Sectra

    - v10.1 P6 - GOOD - black background (default and inverted)

    Cerner Provision

    - Web v4.4.0 (Cedara I-Reach) - BAD - white background

    K-Pacs viewer

    - 1.0 - BAD - white background

    iQ-View

    - 2.0 - BAD - white background

    eFilm

    - v1.53 - GOOD - black background (default and inverted)

    ETIAM LiteBox

    - 3.02 - BAD - white background

    Some additional comments.

    A few folks noticed that even if the image
    was displayed correctly (with a black shutter applied), sometimes
    the thumbnails and icons would not have a black background.

    Some systems have implemented display shutters only for presentation
    states but not for display shutters embedded in images, and others
    have linked their release of display shutters in images to their
    release of presentation state (and hence IHE CPI) capability.

    David

    David Clunie wrote:

    > Hi all
    >
    > Today's crusade is about Display Shutters.
    >
    > I am doing a little survey as a consequence of an incompatibility
    > between two systems (a fluoroscopy modality from one vendor and
    > a PACS from another vendor) reported by a site (see Vince's
    > earlier post to pacs_admin for details if you are interested).
    >
    > How many PACS support application of Display Shutters encoded
    > within images on their workstations or web clients ?
    >
    > I took one of the site's RF images that contains
    > display shutters and has "white" pixels underneath where the
    > X-ray beam has been collimated off, and put it on my web site
    > for anyone who is interested to download and try:
    >
    > http://www.dclunie.com/images/Fluoro...hutter.dcm.zip
    >
    > and let me know what is the result of displaying this normally,
    > and inverting it, together with their PACS make, model and
    > version.
    >
    > This is a de-identified image but from a real product, so it does
    > not test all permutations and combinations of shutters as an
    > exhaustive test set might do, but it does serve to illustrate the
    > practical problem described further below.
    >
    > The DICOM Standard defines a Display Shutter, a means of
    > indicating to the display where the background should be
    > suppressed, regardless of whether or not the image is displayed
    > normally or inverted (i.e., bone and contrast are white or black).
    > It may be circular, rectangular or polygonal or a combination
    > of all three.
    >
    > If this is the means chosen by the modality to suppress the
    > background, and the pixels underneath are "white" in the
    > normal display mode (e.g., if they were collimated off and
    > the detector not exposed), then if the display does not support
    > the shutter then this will result in a mass of white perimeter
    > blinding the user.
    >
    > "Burning in" the shutter as a particular value at the modality
    > (e.g., "black" instead ) doesn't really help much, since if one
    > inverts a white bone or contrast image to get a black bone or
    > contrast image then the background will again go to white.
    >
    > The need to support this feature has been discussed off and
    > on for over a decade of course, as workstations and PACS
    > have had to support fluoroscopy and angiography modalities,
    > though this is not to say that shutters are not also useful
    > for plain projection radiography (CR and digital x-ray) as
    > well.
    >
    > For example, this issue was discussed during the development
    > of the VA/DOD Modality Requirements; see 4.4.4 Pixel Representation
    > in:
    >
    > http://www1.va.gov/imaging/docs/Join..._V_3_0_upd.pdf
    >
    > It specifically mentions that "display workstations
    > should not invert the area outside the shuttered area when
    > the image is inverted", though the document does not actually
    > specify PACS requirements. The initial work on that document
    > probably predates IHE, so it is not as if it is something
    > new that should not have been supported a long time ago.
    >
    > One would have thought that every contemporary PACS would
    > have dealt with this, but I am told that currently deployed
    > versions of some do not (though recent upgrades, at non-trivial
    > cost of course, apparently rectify this problem).
    >
    > Before eviscerating any particular vendor for being
    > particularly lame or belated in this regard, I wanted to find
    > out just how many other systems have this problem, or indeed
    > if users actually perceive it to be a significant problem.
    >
    > David
    >
    > PS. Display shutters are also specified in Presentation States,
    > and they are required to be supported by the standard - in
    > theory any PACS that supports GSPS should support Display
    > Shutters but I am wondering how many actually do in products;
    > the IHE MESA tests for Image Display functionality specifically
    > test for this for actors in the CPI profile. Even if a PACS
    > supports this actor and profile though, and applies display
    > shutters in presentation states, that does not necessarily
    > mean that they will apply those found encoded in images.



+ Reply to Thread