Picker used to sell their IQ line of workstations alongside their CT
and MRs, much the same way GE sells the AW workstations along with
their scanners (Ok, to be fair, they sell the workstations as an
independant product too; but usually at least one gets thrown into the
deal whenever a scanner is sold).

But I digress ... i don't know an IQ workstation would make sense of
those images but suggest you might solict someone who has one of these
workstations (maybe they have one at the scanner where these images
originated?).

Although, looking through the image header, I would anticipate seeing
some private attributes that the Picker software would use to apply a
correction factor to that vertical brightness gradien. There don't
appear to be any private attributes in those images at all. From that I
conclude they've either been removed, possibly through a
de-identification process; or those images look exactly the way they're
supposed to look due to way the images were reconstructed from MR data
or imaging mode used to create them.

I doubt you're going to find any software which is going to display
those particular images in any other way than how they're displaying
now. If you in fact dif de-identified them with some software before
posting them on the web, go back and look at the dicom attributes of
the original images - look for attributes where the first 4 digits of
the attribute tag ends in an odd number. If there aren't any, I don't
think you'll find any software that is going to display them any
differently than what you're seeing now.

Bonjour Andreas,

This does look like a decompression bug. My guess is that they where
originally in a compressed format and at some point where uncompressed
and the program that did this decompression created the problem.

How did you get the images?

Where they like this straight out of the system console? Does the
system have the option of exporting the images in another format
(native maybe?)

Yves



On 23 Feb 2006 09:25:42 -0800, "Andreas Knopke"
wrote:

>Hi Group,
>
>I recently came across some strange MR images from a Picker Polaris
>machine.
>Some of the images have a very inhomogeneous intensity, something like
>a horizontal gradient in brightness and contrast. I was told that there
>are viewers that actually can display these images correctly but all
>viewers I tested (DicomWorks, E-Film, UniPacs, DicomScope, AccuLite)
>showed the same window gradient as my viewer does.
>
>Here are some example images:
>www.k-pacs.net/Image1.dcm
>www.k-pacs.net/Image2.dcm
>
>Does anyone now how to deal with images like this?
>
>regards,
>Andreas
>

Hello Eric and Yves,

thank's for your replies. I got these images from a user who told me
that he can view the images with his workstation at the modality but
not with K-PACS. I don't know if they where compressed and decompressed
but I will investigate further. The strange thing is, that some images
of the series I received seem to be ok. My initial thought was that
this might be an MR artifact.

regards,
Andreas