Storing a modified image

Mats

unread,

Mar 24, 2003, 10:41:25 AM3/24/03

to

Hi.

Assume that you retreive a CR image from a PACS and modify the pixel
data. If you finally want to store the modified CR image (DICOM file)
to the PACS, which Data Elements do you need to update expect the
pixel data?

/Mats

andromeda

unread,

Mar 25, 2003, 12:03:06 PM3/25/03

to

Mats,

this strongly depends on the operation performed on the pixel data.
You should even consider changing the SOPClassUID to Secondary Capture
(SC) which has lots of implications which fields must or may not be
taken from the original image.
What at least should be changed except for the pixel data is:

(0x0008, 0x0070) Manufacturer
(0x0008, 0x1090) Manufacturer model name
(0x0008, 0x0008) ImageType (must not be "Primary")
(0x0008, 0x0032) Image Time
(0x0008, 0x0018) SOP Instance UID

I'm not sure about
(0x0008, 0x0022) Acquisition Date
(0x0008, 0x0032) Acquisition Time

This list may not be exhaustive, and I am not too much an expert in
these things but all the experts seem to be absorbed by the IHE
connect-a-thon, so you now at least have a starting point.

Hope I could help a bit...

andromeda

c97...@cs.umu.se (Mats) wrote in message news:<e56e6970.03032...@posting.google.com>...

Steve Moore

unread,

Mar 27, 2003, 2:08:35 PM3/27/03

to

If this is created by different equipment than the original
equipment, then you need to create a new series for this data.
That implies a new Series Instance UID. You would want to
retain the Study Instance UID.

I'm not sure what to advise about the existence of private elements
in the orignal data set. If you have modified the image, I don't know
if you want to pass private elements on blindly or remove them under
the assumption that they might no longer describe your data.

Steve Moore
moo...@mir.wustl.edu
Mallinckrodt Institute of Radiology

andro...@gmx.net (andromeda) wrote in message news:<df5cf7c4.03032...@posting.google.com>...

David Clunie

unread,

Mar 30, 2003, 6:02:25 AM3/30/03

to

The rule for when to change the SOP Instance UID for an
image is:

"If the pixel data of the derived Image is different from
the pixel data of the source images and this difference is
expected to affect professional interpretation of the image,
the Derived Image shall have a UID different than all the
source images." PS 3.3 C.7.6.1.1.2

It sounds like this is one of those cases when the SOP Instance
UID should be changed.

I would avoid using the Secondary Capture SOP Class for
the derived image if it is at all possible to use CR, since
you will loose support for CR-specific features.

Removing the private elements is a good idea.

Image Type value 1 should be changed from ORIGINAL to DERIVED
as suggested, and the Derivation Description and Source Image
Sequence be set appropriately.

Content Date and Time should probably remain the same,
Acquisition Date and Time should definitely remain the same,
and Instance Creation Date and Time should definitely change.

What to do with the equipment description is now spelled
out in detail in the Contributing Equipment Sequence
description in the SOP Common Module (PS 3.3 C.12.1.1.5)
which was added in CP 270; as Steve points out at the
very least the Series Instance UID needs to be changed.

david

moo...@mir.wustl.edu (Steve Moore) wrote in message news:<a50f0030.03032...@posting.google.com>...

David_F

unread,

Sep 25, 2017, 9:36:22 AM9/25/17

to

I would like to post a question regarding this quote:

"If the pixel data of the derived Image is different from
the pixel data of the source images and this difference is
expected to affect professional interpretation of the image,
the Derived Image shall have a UID different than all the
source images." PS 3.3 C.7.6.1.1.2

For CT images, for instance, when it comes to pixel values the user is typically interested in the Hounsfield User values.
What I encountered already "in the wild" is that there are vendors that consistently modify both the Pixel Data AND the Modality LUT (Rescale Intercept/Slope) of an original CT image such that the User Values don't get modified - and this by keeping the original SOP Instance UID.
So, the semantics of the image didn't change with respect to the User Values.

Is such a (consistent) modification of pixel data and modality LUT without assigning a new SOP Instance UID a legal action to do? If the answer is: Yes, it's legal (though not recommended), would then the wording in quote PS 3.3 C.7.6.1.1.2 for Image modalities with meaningful user values not be better, if it speaks of "user values" rather than "pixel data"?

Many thanks in advance,

David

herman o.

unread,

Sep 26, 2017, 9:34:12 AM9/26/17

to

>
> Is such a (consistent) modification of pixel data and modality LUT without assigning a new SOP Instance UID a legal action to do? If the answer is: Yes, it's legal (though not recommended), would then the wording in quote PS 3.3 C.7.6.1.1.2 for Image modalities with meaningful user values not be better, if it speaks of "user values" rather than "pixel data"?
>
> Many thanks in advance,
>
> David
>

Maybe better to replace "pixel data" with "P-values" as it includes any potential LUT transformations etc.
Herman O.