[Dcmlib] Re: DICOM tags a copier

Emmanuel Olart eolart at theralys.com
Mon Sep 20 21:30:51 CEST 2004 

Salut a tous

Nous avons meme sélectionné moins de fields DICOM que la liste donnée ci
dessous.
Je peux vous extraire ca si ca vous interesse.

Notre but était que les images soient reconnues par e-film qui est
relativement strict sur le format DICOM qu'il accepte de lire.
L'image emptyImage.dcm que je vous avait envoyé est lisible par e-film.

Amicalement
Emmanuel

----- Original Message -----
From: "Jean-Pierre Roux" <Jean-Pierre.Roux at creatis.insa-lyon.fr>
To: <undisclosed-recipients:>
Sent: Monday, September 20, 2004 9:19 PM
Subject: Re: DICOM tags a copier


> > Salut JP,
> >
> > Est-ce que tu peux me confirmer quel tags sont vraiment a copier, de
> > ceux qui sont pas vraiment important. De memoire DicomWorks marquait en
> > rouge ceux qui sont importants, est-ce que c'est dans la norme cette
> > notion de champ 'obligatoire' ?
> >
> > ----
>
> Les champs 'obligatoires' dépendent de la modalité.
> C'est decrit en detail dans la Part 3.3 de DICOM (850 pages) ...
> En particulier, dans la liste que tu proposes ci-dessous
>
> > 0018,0080  Repetition Time: Same as original
> > 0018,0081  Echo Time: Same as original
> > 0018,0082  Inversion Time: Same as original
> > 0018,0083  Number of Averages: Same as original
> > 0018,0084  Imaging Frequency: Same as original
> > 0018,0085  Imaged Nucleus: Same as original
> > 0018,0086  Echo Numbers(s): Same as original
> > 0018,0087  Magnetic Field Strength: Same as original
>
> n'ont de sens que pour des images MR (et ne veulent rien dire pour des
> images CT ou US, ou CR, ou ...)
>
> Theralys, aux dernieres nouvelles, part d'une image
> (gdcmData/emptyImage.dcm) comme modèle pour fabriquer leurs nouvelles
> images.
> son entete (see : PrintHeader emptyImage.dcm )
> est tres peu difference de ce que tu proposes.
> Elle doit bien aller pour 'fabriquer' des images MRI like.
> Si on s'en sert pour une autre modalité, ca doit etre moins terrible
> (des champs obligatoires vont manquer)
> Je n'ai pas fait de test pour savoir si e-film prend ca comme une erreur,
> ou s'il se contente de warnings.
>
>
>
> >
> > Enfin ce qe j'etais en train de pensera un mechisme pour automatiser
> > les accesseurs aux differents champs DICOM. En gros l'utilisateur doit
> > avoir une methode 'SetOrigin' qui prends trois 'float' et qui formate la
> > chaine de caractere correctement. Est-ce que le dico DICOM defini
> > comment doit etre la chaine de caractere ?
>
> Pour une fois, la norme est claire :
> c'est une chaine de caractères, avec le backSlash pour separer les
> composants.
> Dans gdcm, de telles chaines sont generées a coup de std::cout, et
> décomposées avec gdcmUtil::Tokenize
> la VR doit etre "DS" (pour Decimal String); la remarque 'At most 16
> characters' n'a l'air d'etre la que pour le folklore (jamais vu une telle
> chaine de moins de 16 ....)
>
> > Sinon ou est-ce que je peux
> > trouver ca ? Ensuite je peux dire a cmake de parcourir dicomV3.dic pour
> > generer une classe c++ qui automatise ca, pratique qd la norme change...
>
> En ce qui concerne le type des champs, je n'ai jamais vu la norme changer.
> Certains sont dits 'RET' - pour retired-, lorsqu'ils ne sont plus utilises
> dans la norme, mais restent dans le dico Dicom, pour pouvoir continuer a
> lire les anciennes images.
>
> Voila...
>
> JPRx
>
>
> >
> > Merci
> > Mathieu
> >
> > Sinon voila le resultat de ce que je pense etre important :
> >
> > 0002,0002  Media Storage SOP Class UID: Same as original
> >
> > 0002,0003  Media Storage SOP Inst UID: Same as original
> >
> > 0002,0010  Transfer Syntax UID: Same as original
> >
> > 0008,0005  Specific Character Set: Is this same as original or something
> > else?
> >
> > 0008,0008  Image Type: Derived\Secondary
> >
> > 0008,0012  Instance Creation Date: Date when the derived file was
created
> >
> > 0008,0013  Instance Creation Time: Time when the derived file was
created
> >
> > 0008,0014  Instance Creator UID: What goes here?
> >
> > 0008,0016  SOP Class UID: What should this be?
> >
> > 0008,0018  SOP Instance UID: What about this?
> >
> > 0008,0020  Study Date: Same as original
> >
> > 0008,0021  Series Date: Same as original
> >
> > 0008,0022  Acquisition Date: Same as original
> >
> > 0008,0023  Image Date: Same as original
> >
> > 0008,0030  Study Time: Same as original
> >
> > 0008,0031  Series Time: Same as original
> >
> > 0008,0032  Acquisition Time: Same as original
> >
> > 0008,0033  Image Time Same as original
> >
> > 0008,0050  Accession Number: TEST1
> >
> > 0008,0060  Modality: Same as original
> >
> > 0008,0070  Manufacturer: Same as original
> >
> > 0008,0080  Institution Name: Same as original
> >
> > 0008,0090  Referring Physician's Name: Same as original
> >
> > 0008,1010  Station Name: Same as original
> >
> > 0008,1030  Study Description: Same as original
> >
> > 0008,1050  Attending Physician's Name: Same as original
> >
> > 0008,1070  Operator's Name: Same as original
> >
> > 0008,1090  Manufacturer's Model Name: Same as original
> >
> > 0008,1150  Referenced SOP Class UID: Same as original?
> >
> > 0008,1155  Referenced SOP Instance UID: Same as original?
> >
> > 0010,0010  Patient's Name: Same as original
> >
> > 0010,0020  Patient ID: Same as original
> >
> > 0010,0030  Patient's Birth Date: Same as original
> >
> > 0010,0040  Patient's Sex: Same as original
> >
> > 0010,1030  Patient's Weight: Same as original
> >
> > 0010,21B0  Additional Patient History: Same as original
> >
> > 0010,4000  Patient Comments: Same as original
> >
> > 0018,0010  Contrast/Bolus Agent: Same as original
> >
> > 0018,0020  Scanning Sequence: Same as original
> >
> > 0018,0021  Sequence Variant: Same as original
> >
> > 0018,0022  Scan Options: Same as original
> >
> > 0018,0023  MR Acquisition Type: Same as original
> >
> > 0018,0050  Slice Thickness: Same as original
> >
> > 0018,0080  Repetition Time: Same as original
> >
> > 0018,0081  Echo Time: Same as original
> >
> > 0018,0082  Inversion Time: Same as original
> >
> > 0018,0083  Number of Averages: Same as original
> >
> > 0018,0084  Imaging Frequency: Same as original
> >
> > 0018,0085  Imaged Nucleus: Same as original
> >
> > 0018,0086  Echo Numbers(s): Same as original
> >
> > 0018,0087  Magnetic Field Strength: Same as original
> >
> > 0018,0088  Spacing Between Slices: Same as original
> >
> > 0018,0089  Number of Phase Encoding Steps: Same as original
> >
> > 0018,0091  Echo Train Length: Same as original
> >
> > 0018,0093  Percent Sampling: Same as original
> >
> > 0018,0094  Percent Phase Field of View: Same as original
> >
> > 0018,1000  Device Serial Number: Same as original
> >
> > 0018,1020  Software Versions(s): Same as original
> >
> > 0018,1030  Protocol Name: Same as original
> >
> > 0018,1081  Low R-R Value: Same as original
> >
> > 0018,1082  High R-R Value: Same as original
> >
> > 0018,1083  Intervals Acquired: Same as original
> >
> > 0018,1084  Intervals Rejected: Same as original
> >
> > 0018,1088  Heart Rate: Same as original
> >
> > 0018,1250  Receiving Coil: Same as original
> >
> > 0018,1251  Transmitting Coil: Same as original
> >
> > 0018,1312  Phase Encoding Direction: Same as original
> >
> > 0018,1314  Flip Angle: Same as original
> >
> > 0018,5100  Patient Position: Same as original
> >
> > 0020,000D  Study Instance UID:  Same as original?
> >
> > 0020,000E  Series Instance UID: Generated as Synarc root.some series ID?
> > (Can you suggest a specific method?)
> >
> > 0020,0010  Study ID: Same as original?
> >
> > 0020,0011  Series Number: Same as original?
> >
> > 0020,0012  Acquisition Number: Same as original?
> >
> > 0020,0013  Image Number: Blank
> >
> > 0020,0032  Image Position (Patient): - Should be recalculated based on
> > new locations?
> >
> > 0020,0037  Image Orientation (Patient): Should be recalculated based on
> > new locations?
> >
> > 0020,0052  Frame of Reference UID: Blank
> >
> > 0020,0100  Temporal Position Identifier: Same as original
> >
> > 0020,0105  Number of Temporal Positions: Same as original
> >
> > 0020,1040  Position Reference Indicator: Same as original
> >
> > 0020,4000  Image Comments: Same as original
> >
> > 0028,0002  Samples per Pixel: Same as original
> >
> > 0028,0004  Photometric Interpretation: Same as original
> >
> > 0028,0010  Rows: Same as original
> >
> > 0028,0011  Columns: Same as original
> >
> > 0028,0030  Pixel Spacing: Same as original
> >
> > 0028,0100  Bits Allocated: Same as original
> >
> > 0028,0101  Bits Stored: Same as original
> >
> > 0028,0102  High Bit: Same as original
> >
> > 0028,0103  Pixel Representation: Same as original
> >
> > 0028,0120  Pixel Padding Value: Same as original
> >
> > 0028,1050  Window Center: Should this be blank or same as original?
> >
> > 0028,1051  Window Width: Should this be blank or same as original?
> >
> > 0028,1054  Rescale Type: Same as original
> >
> >
>
>

More information about the Dcmlib mailing list