Separate DICOM overlays and SOP Instance UID

[enyaw...@hotmail.com]

Hi:

I have a question about when you need to create a SOP Instance UID for
an image.

I looked at the DICOM standard 2003, part 3 - C.7.6.1.1.3.

Examples of Derived Images that would normally be expected to affect
professional
interpretation and would thus have a new UID include:
a. images resulting from image processing of another image (e.g.
unsharp masking),
b. a multiplanar reformatted CT image,
c. a DSA image derived by subtracting pixel values of one image from
another.
d. an image that has been decompressed after having been compressed
with a lossy
compression algorithm. To ensure that the user has the necessary
information about the
lossy compression, the approximate compression ratio may be included in
Derivation
Description (0008,2111).

But it is not clear to me if a new SOP Instance UID should be created
for the following workflow:

1. PACS sends image to workstation
2. Workstation adds annotations/graphics to image and converts the
annotation to separate DICOM overlay. e.g the graphic data are sent in
Overlay Data Element (60xx,3000) so it did not change the Pixel Data
Element (7FE0,0010).

Since PACS may choose to ignore the image if the SOP Instance UID
already exist in PACS. As a result, the clinical user will not see the
annotations next time he/she views the image from PACS.

Thanks

Wayne Tran

[enyaw...@hotmail.com]

Hi:

I forgot to give step (3).

1. PACS sends image to workstation
2. Workstation adds annotations/graphics to image and converts the
annotation to separate DICOM overlay. e.g the graphic data are sent in
Overlay Data Element (60xx,3000) so it did not change the Pixel Data
Element (7FE0,0010).

3. The workstation sends the image with separate DICOM overlay back to
PACS.

Should the workstation create a new SOP Instance UID for it?

Since PACS may choose to ignore the image if the SOP Instance UID
already exist in PACS. As a result, the clinical user will not see the
annotations next time he/she views the image from PACS.

Thanks

Wayne Tran

[David Clunie]

Hi Wayne

The best thing would be NOT TO DO THIS AT ALL.

It is very bad practice to modify images by adding overlays.

That is what presentation states are for.

But if you do it, there is absolutely no question that you
should assign a new SOP Instance UID - if the overlay contains
information worth preserving then it almost certainly would
'affect professional interpretation' (stupid rule, I know,
it should be less ambiguous than this, but that is history).

Also, pragmatically, as you point out, new UIDs are the only way
to be sure that the PACS doesn't discard it, though there will
then be two "versions" with the same pixel data; further, you will
need to put the new images in a new Series (based on the different
equipment rule), which at least makes it easy to distinguish them.

Finally, many devices will not render overlays in images (whether
they be embedded in (7FE0,0010) or in (60xx,3000)), so presentation
states are much more reliable for this (in the sense that a PACS
either does or does not support them, unambiguously).

All good reasons not to do what you are describing.

In short ... overlays in images suck ... do not use them.

David

[enyaw...@hotmail.com]

Hi Dave:

Thanks for the clarification/explanation.

Wayne Tran