steph.tro@voila.fr wrote:

> Hi,
>
> I'm working on a kind of Viewer which can make operations on the
> DICOM image (e.g. rotations,...). At the present time, when I receive
> DICOM files (CR, CT, PET, MR, ..), I extract pieces of information from
> it and convert image in proprietor format.

Convert images into a proprietary format ?

Heresy !

> I would like to send my files in DICOM format to a server. So, have I
> had to recreate the same Instance UID with a Grayscale Softcopy
> Presentation State (referenced this Instance) or may I send this Image
> (with a rotation) just with conversion jpeg->DICOM ???

Creating a GSPS that refers to the original images is a better bet, but
depends on the viewer downstream to be capable of using them.

Saving multiple versions of an image is a bad thing, since it fills up
the archive, but of course is more likely to be viewable by those
systems that don't support GSPS.

For example, can your own viewer receive and apply GSPS ? It should be
able to if you are going to create them.

The short answer is you need to provide both capabilities.

David

PS. Make sure that any images you generate have different UIDs but refer
back to the original.

PPS. Make sure that any instances of images or GSPS that you create are
valid IODs ... validate them against the standard with one of the many
automated tools available for this.

Hi,
At the present time, my system is an Evidence Creator with the
capacibility to create Secondary Capture. It send this to PACs. I want
to implement an other profile: Key Image Note.I'm able to
Query/Retrieve KIN. I've read that Key Object Selection (Key Image
Note) can referenced to a GSPS. Must I use this GSPS or is it possible
to read KIN without take care about GSPS??

P.S: "Heresy"