Contradiction regarding Structure Set Frames of Reference?

[Andreas Järliden]

Hi,

I am confused about part C.8.8.5.1 of PS 3.3 of the DICOM standard
(2008), and I am hoping that someone can make it a bit clearer for me.

This part talks about the Referenced Frame of Reference Sequence
(3006,0010). First it says:

"The Referenced Frame of Reference Sequence (3006,0010) describes a
set of frames of reference in which **some** or all of the ROIs are
expressed."

If this is read as "a set in which **some** of the ROIs are
expressed", there are ROIs with frames not listed in the sequence. But
that is contradicted by the second part of C.8.8.5.1:

"Since the Referenced Frame of Reference UID (3006,0024) is required
for each ROI, each frame of reference used to express the coordinates
of an ROI shall be listed in the Referenced Frame of Reference
Sequence (3006,0010) once and only once."

which I interpret as meaning that "all ROIs must have there frame of
reference included in (3006,0010)".

This contradicts the first part, so I assume that the first part must
be read as "a set of frames of reference, where in each frame of
reference, some or all of the ROIs are expressed"? But that seems
completely meaningless!

Does anyone have a different interpretation of this? I am reading this
all wrong?

/ Andreas

[bbennett]

Anders,
The referenced frame of reference sequence contains more that just the
frame of reference. It can contain a list of all images which are
used in as a basis for planning( (in the RT_Referenced_Study Sequence
and children thereof). In theory, this is optional, but in practice,
it is required (it is also required in IHE-RO profiles). It may also
contain a transform which translates from this frame of reference to
another. These transforms are seldom implemented, but I have seen
them in the wild. IHE-RO has decided that they should be encoded in
Spatial Registration Objects, and that any RT Structure Set should
have only a single frame of reference (and therefore a single entry in
the Referenced Frame of Reference Sequence.

However, the bottom line is (I believe) that your interpretation is
correct, but your belief that it is meaningless is not. The real
purpose of the entry in the Referenced Frame of Reference Sequence (in
actual practice in the RT community). Is to collect the set of Images
which define the 3D Volume upon which the Structures are Contoured,
and upon which any RT Plans which reference this structure set are
based.

Hope this helps,

Bill Bennett