[Insight-users] Errors while reading DICOM time series (4D) using DicomSeriesReadImageWrite2

[Martin Kavec]

Hi Luis,

and thanks for your message

On Wednesday 11 January 2006 14:43, Luis Ibanez wrote:
> 1) The order of the parameters in DicomSeriesReadImageWrite2
>     seem to be correct. The are:
>
>            1. Directory where your dicom files are
>            2. Filename of the image to be used as output
>            3. Series ID of the DICOM image to read
>
>     Could you please give use more details on how you arrived
>     to the conclusion that parameters (2) and (3) were swapped ?
>

I found the problem, ...., ehm, ... , between keyboard and seat. Apologies!

> 2) It is not surprising that the reader have problems loading
>     your 4D dataset.  The message about slices having repeated
>     positions is to be expected, because, presumably the acquisition
>     protocol is such that for every position in the table, your
>     scanner acquired N time points.
>
>     The DICOM filename generator goes through the list of files
>     in the directory and sort them out based on the table position.
>     Given that you have the N timepoints, the sorting algorithm
>     will find N image with the same table position.
>
>     Please read the following message
>     sent to the list on May 2004:
>
>      http://www.itk.org/pipermail/insight-users/2004-May/008674.html

Thanks for the explanation and the link. I looked at your suggestions to 
Sachin and looked further on the implementation of the GDCMSeriesFileNames. 
However, as far as I can see the sorting based on the three criteria is done 
in GDCMSerieHelper class, so to me it would make sense to implement the 
sorting according to the time on the GDCM level.

However, there is already a method 

void SetUseSeriesDetails( bool useSeriesDetails);

and its documentation says:

 /* Use additional series information such as ProtocolName
   *   and SeriesName to identify when a single SeriesUID contains
   *   multiple 3D volumes - as can occur with perfusion and DTI imaging
   */

If I understand it correctly, this should do what I need, shouldn't it? My 
serie, indeed, contains multiple 3D volumes.

Thanks for your comments.

Martin