Thanks and Regards
Amey.
Display Shutters cover bright, not interesting regions of an image to
avoid glare. If for example an X-Ray is smaller than your lightbox, the
lightbox has some sort of blinds called shutters that cover the unused
area ("make it black").
A presentation display shutter in a DICOM image just states that the
image area underneath the shutter is not of interest and shall be shaded
to black on screen. You do not have a proper presentation if you do not
apply this shutter.
I hope this helps,
Peter
Most importantly, the display shutter if applied properly in
the display application, allows the user to invert the image
(make white black and black white, e.g., when displaying a
digital subtraction angiogram) but keep the shuttered area
black (i.e., not blind the user with a vast expanse of white).
You will find these typically in XA and DX/CR images.
David
Peter B Schmidt wrote:
> Hello Amey,
>
> Display Shutters cover bright, not interesting regions of an image to
> avoid glare. If for example an X-Ray is smaller than your lightbox, the
> lightbox has some sort of blinds called shutters that cover the unused
> area ("make it black").
>
> A presentation display shutter in a DICOM image just states that the
> image area underneath the shutter is not of interest and shall be shaded
> to black on screen. You do not have a proper presentation if you do not
> apply this shutter.
>
> I hope this helps,
>
>
> Peter
>
> Amey wrote:
>> I just now found out a group,elements (0018,1600)-(0018-1624) which
>> tells everything about the display shutters can anybody explain the
>> use of Display shutters.
Hi Amey, they are many types of shutters: rectangular, circular,
polygonal and bitmap shutters. For technical questions please refer C.
7.6.11 Display Shutter Module PS3.3 - you'll find everything there,
even the illustrations :)
I believe you should not apply shutters to pixel values but create
something like "shutter overlays" to give an user ability to switch on
and off the shutters. Algorithm details are left for a programmer and
are not in the scope of this group.
You can take a look at the source in my Java PixelMed toolkit,
which contains:
- a com.pixelmed.dicom.DisplayShutter class for extracting the
information from the DICOM attribute list
- a piece of the pipeline in com.pixelmed.display.SingleImagePanel
that adds to the clip regions of the graphics context
Note that:
- the clipping has to be performed AFTER the image has been
contrast transformed (rescaled and windowed, since the
background should remain a uniform black) but BEFORE
any peripheral window-relative demographics annotations
are applied (since these demographics should remain visible).
- if necessary (as in my code) the clipping is performed
after transforming and selecting the region of the image
to be displayed in the viewport, in which case the image
relative coordinates in the DICOM header need to be
transformed into viewport relative coordinates if the
clipping is performed in the viewport space
- images may contain MULTIPLE shutters, i.e., one each of a
rectangular, circular and polygonal shutter, so you need
to apply ALL of them if present, not just the first one
that you find ... and yes this does happen in real world
images from modalities
See also:
http://java.sun.com/javase/6/docs/api/java/awt/Graphics2D.html#clip%28java.awt.Shape%29
David
Regards
Amey.