[Dcmlib] [Fwd: Re: Unique tag within a dicom file]

[Jean-Pierre Roux]

Mathieu Malaterre wrote:

>
>
> -------- Original Message --------
> Subject: Re: Unique tag within a dicom file
> Date: Tue, 16 Nov 2004 14:57:33 -0500
> From: Andrei Leontiev <aleontievt at hotmail.com>
> Newsgroups: comp.protocols.dicom
> References: <wp9kd.375974$bp1.366155 at twister.nyroc.rr.com>
>
> Hi, Mathieu -
>
> No, there is no "general" key to identify a particular DICOM attribute
> within DICOM Data Set (both DICOM messages and DICOM files contain 
> Data Sets
> augmented by some additional information).

Nous avons donc la confirmation de ce que nous savions déja :
La seule chose qui soit un identifiant, c'est (Group number, Element Number)
-  à l'intérieur du Dicom Dictionary,
- au 'niveau 0' du Header
Si on veut vraiment un identifiant pour *chaque* Dicom Element, il faut 
utiliser l'indentifiant géneralisé créé par Eric, et 'dégouliner' à la 
main chaque 'morceau' :
Le premier donne l'identifiant absolu' au niveau 0;
Le deuxième (s'il existe) donne la position de Seq Item à l'interieur de 
la Sequence
Le troisième donne l'identifiant relatif (à l'interieur de ce Seq Item)
etc.

Quant à l'English Name, c'est encore pire; ce n'est un identifaint nulle 
part.
Les méthodes xxxByName n'ont pas bcp de signification (elles nécessitent 
une consultaion 'visuelle' (par grep) du Dicom Dictionnary par le 
programmeur qui voudrait les utiliser, afin de s'assurer que le nom 
qu'il se propose d'utiliser est bien unique (et n'a pas été modifié 
depuis la derniere version du Dicom Dictionnary)

JPRx

>
> On the top level of Data Set you can only "uniquely identify an attribute
> which may or may not be a sequence. In the latter case, the value of the
> attribute is in fact one or more Data Sets, and you have to parse them
> exactly the same way you did the Top level Data Set. And these Data 
> Sets can
> in turn contain sequences containing Data Sets ad infinitum. DICOM SR
> objects are especially notorious about it.
>
> In other words, you should not consider the structure of DICOM Data 
> Set as
> flat where every element can be accessed directly, via offset, "unique 
> key"
> or any other mechanism, but rather as a tree where you have to 
> navigate from
> one node to another until you reach the leaf node. So, it is not 
> enough to
> know the tag of particular element, but also tags of all sequences it is
> included into, i.e., path from the root to the leaf node.
>
> Andrei
>
> On 11/9/04 3:04 PM, in article 
> wp9kd.375974$bp1.366155 at twister.nyroc.rr.com,
> "Mathieu Malaterre" <mmalaterNOSP at Mnycap.rr.com> wrote:
>
>> Hello,
>>
>> I am a bit confused. I thought that a group + element, would give me a
>> unique key to a Tag. But this isn't true within a sequence. Is there a
>> 'general' key to identify a dicom value, within or not a sequence.
>>
>> Thanks
>> Mathieu
>>
>
>
>
> _______________________________________________
> Dcmlib mailing list
> Dcmlib at creatis.insa-lyon.fr
> http://www.creatis.insa-lyon.fr/mailman/listinfo/dcmlib
>