Importing DICOM files from CD to Philips Jetstream

[Graham Arden]

Hi,

Hope this is the correct group for this question.

I have a series of nuclear medicine SPECT images originally from a
Philips Jetstream 3.0 workstation. As
part of a project I am simulating half time and half angle images, by
removing a proportion of the counts from each image (for one set of
images) and removing alternate frames (for a separate set). I've
written the program to do this using python and the pydicom library.

I've altered several other headers in each file and generated a new
SOP Instance UID and Series Instance UID for each file. So far so
good.

The problem comes when I try and put them back on the Jetstream
workstation. I've burnt them onto a CD and selected "import" from the
Jetstream menu. Selecting "CD containing DICOM images" brings up a
list of the patient names and the DICOM files within them I then
selected import and they appeared to import ok. The problem is that
they do not then appear in the patient browser. I gave them patient
IDs of ZZU0001, ZZU0002...etc in order to force them to the end of the
list but they dont appear anywhere. Searching doesn't bring up
anything.

I checked the import log and they appear to have been transferred,
although interestingly they don't appear in the database log

Any ideas?

Graham

[David Clunie]

In altering the header:

- did you remove all of the private attributes (you may need
to if any are present) ?

- did you change other (non-UID) identifying attributes, like
Study ID and Series Number (you may need to)

- are the instances you generated valid DICOM (or at least,
no less valid than the originals) ... check with dciodvfy

- is the set of instances you are creating internally
consistent (check with dcentvfy)

Also, what tool are you using to make the CD ?

David

[Graham Arden]

Thanks very much for the speedy reply David.

On Jul 2, 8:21 pm, David Clunie <dclu...@dclunie.com> wrote:
> In altering the header:
>
> - did you remove all of the private attributes (you may need
> to if any are present) ?

Haven't yet, although as far as I can see none seem to relate to the
data I have edited.

>
> - did you change other (non-UID) identifying attributes, like
> Study ID and Series Number (you may need to)

Yes, the patient name is set to "Patient_ZZU0001" and the study ID to
"ZZU0001", series number was left as "0"

> - are the instances you generated valid DICOM (or at least,
> no less valid than the originals) ... check with dciodvfy

Just ran through dciodvfy. No errors reported. (Am I correct in
assuming that if the program ends with no error messages everything is
ok?)

> - is the set of instances you are creating internally
> consistent (check with dcentvfy)
>

as above!

> Also, what tool are you using to make the CD ?
>

I used Infrarecorder to burn the CD. All the files are in the root
directory of the CD (I noticed that if I put them in subdirectories
the Jetstream didnt even recognise them as being there.). I've tried
several CDs. All where recognised by a GE Xeleris workstation which
would import the files, but none would import to the Jetstream

Graham

[David Clunie]

Hi Graham

Another thing I can suggest is to import them to the
Xeleris and then try a network transfer to the Jetstream.

Failing that, I guess all you can do is experiment with
removing changing more or different attributes until you
hit the magic combination.

I assume that the Jetsream can re-import its own original
(unchanged) images by the same route (i.e., that import
works at all).

Feel free to send me a set or orignal and modified images
if you want, to see if there is anything obvious wrong
with them (though I do not have a Jetstream to try them
on).

David