[Insight-developers] NRRD Image Format (and tensor estimation thereafter?)

[Gordon Kindlmann]

hello,

Which version of ITK are you using?
The most general handling of NRRD files was added by me prior to the  
2.2 release.

Are you reading the file as a scalar image?
I am guessing this may be the problem.

I believe you can follow the example of:
Insight/Testing/Code/IO/itkNrrdVectorImageReadTest.cxx

which uses a itk::Vector<float, N> PixelType for which the length N  
is known at compile time,

or the example of:
Insight/Testing/Code/Common/itkVectorImageTest.cxx

which uses VectorImage<float, D> image type for which the image  
dimension D is known at compile time, but the length (number of  
values per sample) can be set at run-time, based on a read file.

However, others more knowledgeable than I can chime in with further  
details on the use of VectorImage, also new with ITK 2.2.

Once you have DWIs loaded in, you probably want to do tensor  
estimation.  I am not sure what the current status of ITK-based  
tensor estimation is, specifically, can it work with VectorImages?

Gordon


On Dec 1, 2005, at 10:50 AM, Vincent A. Magnotta wrote:

> I am trying to use the NRRD format to store diffusion weighted  
> images. I
> have been using the information provided on the NA-MIC website:
>
> http://www.na-mic.org/Wiki/index.php/NAMIC_Wiki%3ADTI%3ATensor_format
>
> It appears that the ITK implementation for diffusion weighted images
> does not work properly. From just a first pass analysis, it seems that
> if the NRRD file describes an image that is 4D, but it only has three
> spatial dimensions, the 4th dimension upon loading the images has a  
> size
> of only 1. I have a very small image example that illustrates the
> problem that is attached. When I print out the image after loading  
> it, I
> get the following information:
>
>
> Image (0x84e5bd8)
>   RTTI typeinfo:   itk::Image<short, 4>
>   Reference Count: 2
>   Modified Time: 131
>   Debug: Off
>   Observers:
>     none
>   Source: (none)
>   Source output index:  0
>   Release Data: Off
>   Data Released: False
>   Global Release Data: Off
>   PipelineMTime: 11
>   UpdateMTime: 130
>   LargestPossibleRegion:
>     Dimension: 4
>     Index: [0, 0, 0, 0]
>     Size: [2, 2, 2, 1]
>   BufferedRegion:
>     Dimension: 4
>     Index: [0, 0, 0, 0]
>     Size: [2, 2, 2, 1]
>   RequestedRegion:
>     Dimension: 4
>     Index: [0, 0, 0, 0]
>     Size: [2, 2, 2, 1]
>   Spacing: [2, 2.5, 2, 1]
>   Origin: [0, 0, 0, 0]
>   Direction:
> -1 0 0 0
> 0 -1 0 0
> 0 0 1 0
> 0 0 0 1
>
> I will look into this some more over the next couple of days, but I
> thought that I would post it here as well. I am just getting familiar
> with the format, so if anyone else has suggestions they would be
> helpful.
>
> Vince
>
>
>
> -- 
> Assistant Professor
> Department of Radiology
> 0453-D JCP
> 200 Hawkins Drive
> Iowa City, IA 52242
> E-mail: vincent-magnotta at uiowa.edu
> Phone: 319-356-8255
> Fax: 319-353-6275
> Website: http://www.radiology.uiowa.edu
> <tmp.img>
> <tmp.nhdr>
> _______________________________________________
> Insight-developers mailing list
> Insight-developers at itk.org
> http://www.itk.org/mailman/listinfo/insight-developers