Understanding how to display DICOM RT-Dose
johnw...@gmail.com
I am trying to better understand the concepts behind DICOM RT-Dose
objects, and how they should be displayed by viewing applications
supporting DICOM RT objects.
What I understood until now (correct me if I am wrong) is that doses
can be expressed in terms of:
1) dose points or isodose curves: in this case, the elements which
should be displayed by a RT viewer application are described in the
DICOM dataset through the ROI Contour module, which basically
describes points, lines or polygons as a set of 3D coordinates in the
3D space of the patient.
2) does grid: in this second case, the dose values are described
through the Pixel Data data element. Here, pixel values actually
represent dose values, correct? A multi-frame Pixel Data attribute may
be present, and in this case each frame represent dose values for a
given "plane / slice", as specified by the "(300C,000C) - Grid Frame
Offset Vector" data element.
Is my understanding correct until now?
Ok, now I have a question:
I have some sample studies where multiple "RT-Dose" DICOM files are
present, each one having multiple frames in Pixel Data. Considering
all these RT-Dose DICOM files, I have several RT dose "pixel data"
frames (coming from different files) which apply to a given CT slice
(i.e., which belong to the same plane as the CT slice). Should I SUM
the dose values coming from each RT dose frame which apply to a given
CT slice in order to obtain the actual dose value which should be
applied to every pixel of my CT slice? Or how should I combine values
coming from different RT dose frames which apply to the same CT slice
(i.e., which have the same Grid Frame Offset value)?
In particular, I am considering the issue from the point of view of
the viewing application: I have seen applications which show the RT
dose which applies to each CT slice as a semi-transparent overlay area
varying from blue for lower-dose zones to red for higher-dose zones
(IsoShades). So in this case I guess the viewing application should
compute the "overall" dose which applies to a given CT slice. How are
the several RT-dose frames related to a given CT slice's plane
combined?
Any clarification would be greatly appreciated.
Regards,
John.
Singani
Let me add in my views....
Dose can contains the Grid doses which stroed in pixeldata (ImagPixel
Module Attribute) , you have to multiply this each pixel value with
DoseGridScaling(0x3004000E) then you will retrieves the Griddoses
which is in dose units specified in DoseUnits attributes.
IsoDoseCurves or IsoDosePoints can be stored in ROICOntour and
structureset , roi dose module points of curves which having same dose
across the curves....
To view this dose , you have to get the current CT image plane cut
this plane in dose volume which constructed based on pixeldata and
dose grid scaling display the overlay on ctimage.. At a time one dose
only you have to display on top of ct volume..
Regards,
KK
bbennett
An RT-Dose represents a calculated dose. It references an RT-Plan,
which, in turn, references an RT-Structure Set, which, in turn,
references a set of images (usually CT's). Calculation of Dose is
based primarily on the RT-Plan, but may involve use of data in the RT-
Structure Set, and even the CT's. It also involves models of
Radiation Source which are outside of DICOM.
The Dose, the RT-Structure Set, and the CT's are almost always in the
same frame of reference, so (for display purposes) you can shortcut
all of these references if you know that the Dose and the CT's are in
the same frame of reference. The Dose represents the calculated dose
for a given "summation type" and a particular plan. A summation type
can be "PLAN", "FRACTION", "BEAM", "BRACHY", or "CONTROL_POINT" (In my
experience, I have only seen "PLAN", "FRACTION", and "BEAM"). A
"PLAN" is the most inclusive summation type. (Currently an RT-Dose
cannot represent a dose for more than a single plan).
Most often the summation type is "PLAN", i.e. the RT-Dose represents
the Dose for the entire plan. It is wrong to sum the doses for
multiple plans. Summing the doses for multiple beams in the same
plan, on the other hand makes more sense.
I don't know the particular characteristics of the RT-Dose and CT's
that you have, but the planes of grid based RT-Dose are often not co-
incident with the CT's upon which you would display the dose, so you
often have to interpolate between the dose planes.
Generally, questions of dose summation are best addressed to a medical
physicist, not a DICOM expert, as the really relevant questions relate
the the underlying physics, and not to DICOM encoding of dose.
HTH,
Bill B
johnw...@gmail.com
The RT-Dose DICOM files I am dealing with have "summation type" =
"BEAM", and there are multiple RT-Dose files (i.e., multiple RT-Dose
grid pixel data) referring to a given CT plane (image), and referring
to a single RT-Plan DICOM object. It looks like there is one RT-Dose
file for each beam described in the RT-Plan.
In this case, do you believe it would be appropriate to sum all RT-
Dose grid pixel values (after having multiplied them by the
"DoseGridScaling" value) related to each beam in order to build up the
"overall" dose image to be displayed as an overlay to the CT image?
Finally, a more general question: can you recommend a documentation
source where I could find some more information about RT? I agree with
you that some of my questions are more "medical physicist's" questions
than DICOM-related questions, but I don't know where to find
appropriate information.
For instance, now I would also like to understand something more about
how to display the various beams described by the RT-Plan DICOM object
over the CT image(s) they refer to, and I guess I would need some
background information about RT Plan and how the radiation beams are
setup in order to better understand the meaning of the various values
stored in the RT Plan DICOM object...
Any suggestion would be highly appreciated.
Best regards,
John.