Hi Felix,

Most filters are defined generic. For example, they are using a specific NxN 
matrix for filtering, where N is the image dimension. This way you can e.g. 
perform edge detection on 1D, 2D or 3D images. Usually it would not make much 
sense to perform something like temporal edge detection in 4D images.

In most cases, you probably would like to perform a 3D-filter onto each 
timestep-3D-volume. You can do this by extracting the single timesteps as of a 
4D image as 3D images, perform the filter onto these and eventually put them 
back together to a 4D image.

Pseudo-code example:

   // input is 4D mitk Image
   mitk::Image::Pointer output = input->Clone();

   for (int t=0; t<input->GetTimesteps(); t++)
   {
       mitk::ImageTimeSelector::Pointer timeSelector = 
mitk::ImageTimeSelector::New();
       timeSelector->SetInput( input );
       timeSelector->SetTimeNr( t );
       timeSelector->Update();

       mitk::MyFilter::Pointer filter = mitk::MyFilter::New();
       filter->SetInput(timeSelector->GetOutput());
       filter->Update();

       output->SetVolume( filter->GetOutput(), t);
   }


Of course we are happy about contributions! However, it would be nice if you 
could consider the following rules:
General Rules for 4D filtering:

-          If filters are just not accepting 4D images as input or crash in the 
processing, this is of course a bug.

-          The filters should PRIMARY perform generic processing (e.g. 4D edge 
detection, even if the developer wants this for each timestep separately).

-          If the filters generic processing is possible for 4D images, but the 
developer wants the processing be done for each timestep separately, you could 
provide an additional parameter for this behavior. (This needs to be documented 
clearly in the filter class)

-          If due to the filters nature the generic processing for 4D images is 
just not possible, it should perform 3D filtering on each timestep. (this 
should be documented clearly in the filter class).





Regards,

Bastian


Von: Félix C. Morency [mailto:[email protected]]
Gesendet: Montag, 10. Dezember 2012 17:42
An: [email protected]
Betreff: [mitk-users] 4D filters

Hi list,

I've been investigating the 4D support of MITK and noticed that many filters 
behave badly or simply doesn't work in the presence of 4D data. Such examples 
are the crop filter, the region growing filter and the segmentation filter. I 
would like to know if anyone is working on better support of 4D data/filters? 
Does anyone had success with 4D data processing?

This is required for the project I'm working on so me and my colleague will 
probably make contributions regarding this. Please let me know if you are 
interested in this.

Regards,
--
Félix C. Morency, M.Sc.
Plateforme d’analyse et de visualisation d’images
Centre Hospitalier Universitaire de Sherbrooke
Centre de recherche clinique Étienne-Le Bel
Local Z5-3031 | 819.346.1110 ext 16634
------------------------------------------------------------------------------
LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial
Remotely access PCs and mobile devices and provide instant support
Improve your efficiency, and focus on delivering more value-add services
Discover what IT Professionals Know. Rescue delivers
http://p.sf.net/sfu/logmein_12329d2d
_______________________________________________
mitk-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mitk-users

Reply via email to