Hi Daniel, I think I managed to get the image on which I projected and filled the contour of the mitk::contourTool, but know I have to write back this "reoriented slice" on the original working data. This is normally done by the OverwriteSliceImageFilter but It's not possible to use the same approach in this case.
Do you have any idea on how to do that? I had a look at some mappers (imageDataMapper and geometrye2dDataMapper3d) but I'm stacked and I don't know how I could do that Could you give me some hints please? thank you very much chiara ________________________________________ From: Daniel Maleike [[email protected]] Sent: Friday, June 26, 2009 11:51 AM To: [email protected] Subject: Re: [mitk-users] 2d segmentation tool on reoriented slices On Thu, 2009-06-25 at 14:51 +0200, Chiara Riccobene wrote: > Hi mitk-users! > > In my lab we're planning to implement a manual segmentation on reoriented > slices, we would like to know if you're are also developing something similar > or if you plan to do it in the future. > > We think that we could implement something similar to mitk::Seg2dTool where > instead of looking for the pixel in the current slice we will look for the > pixel in the working volume image (working data), the most complicated part I > think would be the filling of this 3d contour in the volume image. > > Do you have any idea of how to do that? Did you already implement something > similar? > > thank you very much in advance for your attention > regards > > Chiara Riccobene Dear Chiara, the tool you describe would be very valuable indeed. Especially colleagues working with MR images are asking for something like that. About the implementation idea, I would prefer to change mitk::SegTool2D in a way that it can cope with reoriented slices. I know that the current implementation is missing a concept for this, but we could change that. The simple part in implementation would be to extract a 2D slice from the 3D volume -- the 2D rendering does that and VTK provides a flexible filter for reslicing. The harder part will be to "write back" 2D slices to a 3D volume. When this is done for a single slice, I see no problem (only a minor definition problem of how thick the slice is). However, if you do that for consecutive (reoriented) slices, I an not really sure how you can guarantee that you do not generate single pixel holes in structures that you wanted to segment as a block. One solution could be to define slice thickness larger than slice offset. To answer your question clearly: nobody here is currently working on this issue and I am not aware of plans to do so. But as I mentioned, there is definitely a demand for this feature, so I would support integrating changes towards this direction. If you have suggestions for required changes, please feel free to send them in and start a discussion. Would you be willing and able to contribute these changes to MITK? Kind regards Daniel -- Dipl.-Inform. Med. Daniel Maleike Phone: +49 6221 42 2326 Deutsches Krebsforschungszentrum Im Neuenheimer Feld 280 Medical and Biological Informatics (E130) 69120 Heidelberg ------------------------------------------------------------------------------ _______________________________________________ mitk-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/mitk-users ------------------------------------------------------------------------------ _______________________________________________ mitk-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/mitk-users
