[vtkusers] dicom into 3D

Sebastien BARRE seb-ml-vtk at barre.nom.fr
Sat Mar 10 22:10:02 EST 2001

At 11/03/2001 10:24, Audi wrote:

> > Did anybody ever tried to load DICOM samples provided in the sebastien
>Barre website?

I guess you are referring to that page :
and the file : CT-MONO2-16-ankle

> > I tried to load DICOM sample about ankle and it only able to display 2D
> > image eventhough I use volume rendering.
> > Are you able to render the DICOM samples after you convert to RAW and
> > display it as 3D that can rotated such that can be seen all side of the 
> image?

Oh I see the problem :), there is some kind of misunderstanding :) That 
file does not hold any "3D object", it's just a *2D* screen-capture of a 3D 
reconstruction that was performed at a hospital, somewhere, somehow, 
sometime (I do not know much about that picture, and I've anonymized it 

Look at its tags :
                    Image Type (0008,0008)  1-n  CS [DERIVED\SECONDARY\3D]
that kind of file is a so-called "secondary capture", a "A general catch 
all Modality that is used for images that did not originate as a DICOM 
image. Image that were stored as TIFF or BMP that are converted to 
DICOM  are given this Modality.".
Plain 2D.

> > I really stuck, I suppose to render that image in 3D, but I always get 2D.

Definitely. There is hardly any chance that you will ever get a 3D volume 
rendering from a single-frame DICOM file. You will succeed by using a set 
of contiguous 2D images, i.e. a *series* of medical file (an acquisition). 
Or a multiple-frame DICOM file, although I've not seen any stack organized 
that way at the moment (multiple-frame file are most of the time used for 
cine-mode and so on).

> > If anybody can display it as 3D, can give me advise how to do it?

If somebody has displayed it as 3D, it deserves my admiration, and should 
send me a postcard :)

Anyway, if you want to try "volume rendering", download the VTK data files, 
and use the frog, fullhead, headsq dataset.

More information about the vtkusers mailing list