[Insight-users] Modality LUT transform with gdcmIO

[Mathieu Malaterre]

Jesús Spí­nola wrote:
>     This is fixed in ITK CVS, thus it should be available in the coming
>     release.
> 
>     Thanks again for your bug report,
>     -M
> 
> 
> That's great!
> About the original pixel range of the  xlut_01.dcm test image maybe it's 
> an errata when I told it had a stored pixel range of 128..255 because I 
> extracted that info from the "IHE Year 5 Test Plan for Image Displays" 
> document. The other info is retrieved with the dcmtk library command 
> line tools ( "dcmdump" to be more exactly ). Which tool do you recommend 
> me to extract the original stored pixel range?

Oh I see. It was encoded in the image comment :)

> I also have another question about the Photometric Interpretation. When 
> the image is of type MONOCHROME1 ( higher values are darker, lower 
> values are brighter ) does gdcm reverse the data to be MONOCHROME2 when 
> it's loaded to itk or it does nothing with the Photometric Interpretation?

correct. Basically you can assume that GDCM will always do the 
conversion to the ITK model for you.

> And I ask again, when a modality lut is present ( tag 0x0028,0x3000 ), 
> does gdcm apply it to the image?

Well that's simply a bug. We have had some discussion in the past to 
support that, but we never did. So gdcm (itself) cannot handle those 
LUT. Since this is a fairly big task, could you please enter a bug in 
the bug tracker so we don't loose reference to this issue:

    http://itk.org/Bug

As a side note, you understand those datasets are for testing only. Are 
you actually dealing with such images in real life ?

Thanks,
-Mathieu