Measurements in CR, DR and US studies. - DICOM

This is a discussion on Measurements in CR, DR and US studies. - DICOM ; Hi, I read few of the old messages on this topic and I am still cofused regarding this. What parameter, Pixel Spacing or image pixel spacing should be used while calculating the length? Also this technique will work on all ...

+ Reply to Thread
Results 1 to 3 of 3

Thread: Measurements in CR, DR and US studies.

  1. Measurements in CR, DR and US studies.

    Hi,

    I read few of the old messages on this topic and I am still cofused
    regarding this.

    What parameter, Pixel Spacing or image pixel spacing should be used
    while calculating the length?

    Also this technique will work on all (XA,CR,DR,US) images or there are
    differenret methods for different images.

    I downloaded Test images of date 20050624 for Pixel spacing and they
    say image pixel spacing should be used. Is this correct?

    Also does this parameter exist in ultrasound images or not?

    Regards,
    Prashant


  2. Re: Measurements in CR, DR and US studies.

    Hi Prashant

    You need to review CP 586 which addresses this specific concern. It
    is already incorporated in the 2007 edition, but the rationale may
    be of use to you.

    The bottom line is that when both attributes are present and have
    different values, you should alert the user rather than risk making
    an incorrect assumption ... there are two potential measurement
    results (one "calibrated" perhaps, and one not), and only the
    user can make the appropriate distinction based on their
    knowledge of the equipment involved - this is especially true
    since the "calibrated" value may not be "better" than the
    uncalibrated value, because any a priori calibration involves
    assumptions that may not be true (or in some cases may not even be
    known to the operator or the service engineer or the vendor, it
    seems). The Imager Pixel Spacing value is the one that is supposed
    to most closely approximate what would happen if the user measured
    distance on a film from a film-screen cassette exposure.

    The better and safer way to make absolute (as opposed to relative)
    measurements on projection radiographs is to use a fiducial of
    known size close to the central ray in the same plane as the
    object being measured, and your software should allow for this
    type of "calibration" to be performed at display time, but you
    probably already know this.

    Ultrasound is completely different ... they use the Region
    Calibration capability described in the standard to specify
    the size of pixels in different parts of the image. Some
    ultrasound vendors may send Pixel Spacing values (instead,
    or as well as Region Calibration) but these do not have a
    standard meaning since they are not in the standard Ultrasound
    IOD.

    It is convenient, but potentially dangerous, to assume that
    measurements can be made on ultrasound images using Pixel
    Spacing, if present.

    David

    Prash wrote:
    > Hi,
    >
    > I read few of the old messages on this topic and I am still cofused
    > regarding this.
    >
    > What parameter, Pixel Spacing or image pixel spacing should be used
    > while calculating the length?
    >
    > Also this technique will work on all (XA,CR,DR,US) images or there are
    > differenret methods for different images.
    >
    > I downloaded Test images of date 20050624 for Pixel spacing and they
    > say image pixel spacing should be used. Is this correct?
    >
    > Also does this parameter exist in ultrasound images or not?
    >
    > Regards,
    > Prashant
    >


  3. Re: Measurements in CR, DR and US studies.

    Luigipb, David,

    Thanks for the reply. I read the CP 586. What I understood is image
    pixel spacing is important than pixel spacing. Both the values should
    be same in most of the cases. In calibrated images they can differ and
    pixel spacing indicates the calibrated value while image pixel spacing
    uncelebrated. In such cases we should alert user to select appropriate
    method.

    This sounds good but I really doubt if this will really help or not.
    Will users understand all these technical terms (calibrated,
    uncalibrated, pixel spacing, linear spacing) and make judgment
    accordingly? Are users in Asian or African subcontinents are so tech
    savvy to understand all this?

    Regarding the objects of known size in the image, the selection of
    image used for calibration will matter all the measurements. Means it
    again dependent on external factors.

    Regards,
    Prashant


+ Reply to Thread