[vtkusers] vtk4.4 DICOM

[Day, Robert]

> -----Original Message-----
Jeremy Winston [mailto:jbw at ieee.org] wrote:
> You have to be careful--not all companies' DICOM
> implementations are created equal.
> 
> In order to use vtkDICOMImageReader on a stack of
> images generated by a Siemens machine in their
> Somatom5 format, I first had to rename each slice
> file using ezDICOM
> http://www.psychology.nottingham.ac.uk/staff/cr1/ezdicom.html
> since the Siemens software seems to pick file names
> out of the air -- i.e., they're not ordered in any
> obvious way.
> 

This is, of course, the correct way for a DICOM implementation to work.
The standard explicitly states that file names are not to be parsed for any
information about the data set.
That should always be parsed out of the relevant tags in the file.

Of course, I always number mine by slice number anyway ;-)
dicom2 <www.barre.nom.fr/medical/dicom2/> is the canonical way to do this, IMHO.

Rob.
--
Robert Day            ph  +61 8 9224 3227
Project Bioengineer   fax +61 8 9224 1138
Royal Perth Hospital  robert.day at health.wa.gov.au