[vtkusers] VolumeReader for compressed data

[Sebastien BARRE]

Hi Francis, Marc

I guess I did write a FAQ entry about that, didn't I :)) ?
http://public.kitware.com/cgi-bin/vtkfaq?req=all#3.6

At 16:50 08/08/00 +0200, M.Schrijver at el.utwente.nl a écrit:

>probably the best way of reading Dicom files is by using Sebastien
>Barre's dicom2 tool
>(http://www.hds.utc.fr/~barre/medical/dicom2/index.html) to convert it
>to a set of 'images'. This works perfectly for (X-ray) images, I'm only
>not sure if it works for volumes as well. Just give it a try.

Thanks for the link. Such beast "would" have been the "best" way if it 
would have been able to read compressed syntax, but sadly, it's *not* (see 
the home page :). It's not obvious to write a DICOM reader, and even harder 
to implement DICOM compressed syntax support, believe me :)

My tools works on uncompressed (native) DICOM syntax at the moment : I'm 
able to read DICOM files (single or multi frame), mass/batch-convert them 
to RAW files, then feed them to VTK to perform 3D reconstruction. Fairly 
standard. The whole stuff is described here :

http://www.hds.utc.fr/~barre/medical/dicom2/how-to.html#Convert to raw

>Then you'll have to find out the exact type of compression used on the
>data, as specified by the 'transfer syntax' in the Dicom file. I
>remember using the same tool (dicom2) to get that kind of information as
>well, but don't ask me how at the moment....

dicom2 -t1 yourfile | grep (0002,0010)

>  If it turns out to be
>"1.2.840.10008.1.2.4.70", indicating it is "JPEG lossless,
>non-hierarchical, first-order prediction (process 14, selection value
>1)" (!) encoded data, well, in that case

... in that case, my tool can not help you :(

Syntax are listed here :
http://www.hds.utc.fr/~barre/medical/samples/index.html#transfer-syntax

>I might be able to help you.
>Some time ago I wrote a subclass of vtkImageReader (which reads volumes
>consisting of multiple slices/images very well) that can do the
>reconstruction for you.

You mean, it is *able* to read compressed syntax ? With JPEG support for 
example ? Would be nice.