[Insight-developers] DICOM 0008:0008

[Mathieu Malaterre]

Hi David,

On Sat, Apr 19, 2008 at 2:23 PM, David Clunie <dclunie at dclunie.com> wrote:
> Hi Mathieu
>
>
>  Mathieu Malaterre wrote:
>
> > On Wed, Apr 9, 2008 at 2:22 PM, Miller, James V (GE, Research)
> > <millerjv at crd.ge.com> wrote:
> >
> > >
> > >
> > > The DICOM Image Type (0008:0008) is usually something like:
> > > ORIGINAL\PRIMARY\AXIAL when DICOM files are coming off a scanner.  ITK
> > > generated files have the Image Type as DERIVED\PRIMARY.  Should the
> Image
> > > Type be DERIVED\PRIMARY\AXIAL when appropriate?
> > >
> >
> > Hi Jim,
> >
> >  I am guessing you are talking about a CT Image Storage DICOM object,
> right ?
> >  AFAIK Value #3 of Image Type is optional, so it should not be needed
> > (only the first 2 are required).
> >
>
>  Image Type (0008,0008) Value 3 is not optional for CT images - it is Type
> 1;
>  see PS 3.3 C.8.2.1.1.1.

Quoting 'C.7.6.1.1.2 Image Type':

The Image Type attribute is multi-valued and shall be provided in the
following manner:
a. Value 1 shall identify the Pixel Data Characteristics; Enumerated
Values for the Pixel
Data Characteristics are:
ORIGINAL identifies an Original Image
DERIVED identifies a Derived Image
b. Value 2 shall identify the Patient Examination Characteristics;
Enumerated Values for the
Patient Examination Characteristics are:
PRIMARY identifies a Primary Image
SECONDARY identifies a Secondary Image
c. Value 3 shall identify any Image IOD specific specialization (optional)
d. Other Values which are implementation specific (optional)

I believe the 'optional' term for value #3 and up is clearly misleading.

>  So if you discard or do not replace value 3, your result will be
> non-compliant.

After rereading the specifics of Image Type for CT Image Storage, I
understand since value #3 is a Defined Term, thanks for correction.

>  This error should be detected if you run a validator (like my dciodvfy) on
>  the output, which should be a routine step for anyone writing DICOM files.

dciodvfy remains one the basis tool (and dcmdump from dcmtk) used in
our nightly testing for GDCM, thanks again for this great tool !

>  Note also that AXIAL does not mean "transverse", it means any
> cross-sectional
>  image, and specifically "not a localizer", so always specifying AXIAL when
>  the source was AXIAL should be quite safe.

Regards,
-- 
Mathieu