DICOM GSDF question (PS3.14-2004 page 9) - DICOM

This is a discussion on DICOM GSDF question (PS3.14-2004 page 9) - DICOM ; After reading PS3.14-2004, following questions come up: 1. Figure 6.1 shows "Modality -> VOI -> Polarity -> Presentation -> Standard display system". Modality here means data from "Modality LUT IE" in an IOD? 2. "VOI" here means transformation based on ...

+ Reply to Thread
Results 1 to 6 of 6

Thread: DICOM GSDF question (PS3.14-2004 page 9)

  1. DICOM GSDF question (PS3.14-2004 page 9)

    After reading PS3.14-2004, following questions come up:

    1. Figure 6.1 shows "Modality -> VOI -> Polarity -> Presentation ->
    Standard display system". Modality here means data from "Modality LUT
    IE" in an IOD?

    2. "VOI" here means transformation based on "VOI LUT IE"?

    3. "Polarity" here means transformation based on what?

    4. "Presentation" here means transformation based on "PRESENTATION STATE
    IE"?

    5. Are things like "Curve IE", "Palette color lut" used in this process?
    Any other information (IOD modules) used in this process but not
    mentioned above?

    6. After presentation lut transformation, we will get "both device
    independent and conceptually perceptually linear" data. Is this data
    linear "relative to" what? the "strength" of x-ray signal? density of
    the object/person being tested? Let's just use CR and DX as examples here.

    7. The part I do not understand is why not do window width/center kind
    of "linear" transformation (considered as part of VOI LUT) AFTER
    presentation transformation? In other words, shouldn't the device
    independent and conceptually perceptually linear P-Value be the ideal
    space to run any linear transformation based upon?

    If you run VOI before "Polarity -> Presentation", your linear VOI
    transformation may no longer be linear in P-value space. What is the point?

  2. Re: DICOM GSDF question (PS3.14-2004 page 9)

    8. how does psudo-color palette fit into the whole picture. Is it a
    presentation LUT?

    9. some of the viewers out there allow additional LUT set by the user,
    some are non-linear such as log-based. Are those considered
    presentation LUT? Should they be applied after original presentation
    LUT, before VOI, or after VOI, before polarity? Or else?
    Thanks in advance.


  3. Re: DICOM GSDF question (PS3.14-2004 page 9)

    another stupid question here:

    10. Considering the traditional negative/film, is it equivalent to the
    signal after modality LUT transformation? VOI? Presentation LUT? Or
    after display system (including GSDF applied). If the last is true,
    what process guarantees the linearity of films and what process applies
    the right GSDF to it?

    It appears to me that making the LCD display "the same" as traditional
    film is one of the goals of this DICOM image rendering process.
    However, are you sure that the file we used previously were good
    approximations of things we want to get in here? Coming from a film
    viewing background, how can a radiology judge if a LCD "looks right"?


  4. Re: DICOM GSDF question (PS3.14-2004 page 9)

    Change the world "file" to be "film" in previous post:

    "However, are you sure that the file we used previously were good
    approximations of things we want to get in here? "


  5. Re: DICOM GSDF question (PS3.14-2004 page 9)

    You might want to take a look at the description of the grayscale
    pipeline for softcopy in PS 3.4 Annex N, in addition to the detailed
    description of the cascading behavior of the Modality and VOI LUT
    steps described for those Modules in PS 3.3 C.11.

    Those sections should answer most of your questions I think.

    David

    nu pacs wrote:

    > After reading PS3.14-2004, following questions come up:
    >
    > 1. Figure 6.1 shows "Modality -> VOI -> Polarity -> Presentation ->
    > Standard display system". Modality here means data from "Modality LUT
    > IE" in an IOD?
    >
    > 2. "VOI" here means transformation based on "VOI LUT IE"?
    >
    > 3. "Polarity" here means transformation based on what?
    >
    > 4. "Presentation" here means transformation based on "PRESENTATION STATE
    > IE"?
    >
    > 5. Are things like "Curve IE", "Palette color lut" used in this process?
    > Any other information (IOD modules) used in this process but not
    > mentioned above?
    >
    > 6. After presentation lut transformation, we will get "both device
    > independent and conceptually perceptually linear" data. Is this data
    > linear "relative to" what? the "strength" of x-ray signal? density of
    > the object/person being tested? Let's just use CR and DX as examples here.
    >
    > 7. The part I do not understand is why not do window width/center kind
    > of "linear" transformation (considered as part of VOI LUT) AFTER
    > presentation transformation? In other words, shouldn't the device
    > independent and conceptually perceptually linear P-Value be the ideal
    > space to run any linear transformation based upon?
    >
    > If you run VOI before "Polarity -> Presentation", your linear VOI
    > transformation may no longer be linear in P-value space. What is the point?



  6. Re: DICOM GSDF question (PS3.14-2004 page 9)

    As far as pseudo-color is concerned, you might find the new pipeline
    described in the proposed Supplement 100 of help.

    See the public comment text at:

    ftp://medical.nema.org/medical/dicom.../sup100_pc.pdf

    David

    chinapacs2003@gmail.com wrote:

    > 8. how does psudo-color palette fit into the whole picture. Is it a
    > presentation LUT?
    >
    > 9. some of the viewers out there allow additional LUT set by the user,
    > some are non-linear such as log-based. Are those considered
    > presentation LUT? Should they be applied after original presentation
    > LUT, before VOI, or after VOI, before polarity? Or else?
    > Thanks in advance.
    >



+ Reply to Thread