[Dcmlib] [Fwd: Re: dicom3tools_1.00.snapshot.20050804.helgray.tar.bz2] David Clunnie's anwer

[Jean-Pierre Roux]

Mathieu Malaterre wrote:

> For refs:
>
> -------- Original Message --------
> Subject: Re: dicom3tools_1.00.snapshot.20050804.helgray.tar.bz2


--------------------------------------------------------------------------------------------------------

> However DICOM PS 3-5 7.8.1 a) states that those (gggg-0010->00FF where
> gggg is odd) attributes have to be LO.

Yes, but that has nothing to do with the way my dictionary is
constructed.

> Here is a concrete question :
> If we are decoding a dataset and we encounter a tag like (0009,0012) :
> - when the dataset is encoded in EXplicit with a VR different of LO, we
> may assume its VR is invalid

Yes.

> - when the dataset is encoded in IMplicit we can assume that its VR is
> LO

Yes.

----------------------------------------------------------------------------------------------------------

In gdcm, we choosed to trust the 'EXplicit' VR, better than the Dicom 
Dictionnary
 -we assumed the 'writter' knows what he's doing (hope so), and we want 
to read any image-

What we can do easyly, is to force to VR=LO any odd dicom tag whose 
number is 10<   <FF

Any objection ?
JPRx

>
> Hello David
>
> I don't really know how your dictionary is encoded, but I understand
> that it may use some masks, and that you may use them internally in
> your toolkit.
>
> However DICOM PS 3-5 7.8.1 a) states that those (gggg-0010->00FF where
> gggg is odd) attributes have to be LO.
>
> Here is a concrete question :
> If we are decoding a dataset and we encounter a tag like (0009,0012) :
> - when the dataset is encoded in EXplicit with a VR different of LO, we
> may assume its VR is invalid
> - when the dataset is encoded in IMplicit we can assume that its VR is
> LO
> Then all tags found at the same depth level, and between (0009,1200)
> and (0009,12FF) belong to this group.
>
> Am I wrong ?
> Have a nice day !
>
> Gilles
>
> _______________________________________________
> Dcmlib mailing list
> Dcmlib at creatis.insa-lyon.fr
> http://www.creatis.insa-lyon.fr/mailman/listinfo/dcmlib
>