[Insight-users] RE: GDCM image size reading

Li, George (NIH/NCI) ligeorge at mail.nih.gov
Fri May 20 12:09:05 EDT 2005


Mathieu:

Thanks for the note. I have identified a portion of the problem, 
which is the inconsistency in the image sizes among multiple images 
that I intend to combine them together in a multi-filed voxel volume. 
The size difference will cause problem and can be solved with some
data manipulation. I should have verified the image data first.

However, there seem other problems and I am trying to identify them.
I hope that they do not fall into the case that you suspected in
terms of the "flag" inconsistency and consequently misusage of the
system libraries.

Regards,

George


-----Original Message-----
From: Mathieu Malaterre [mailto:mathieu.malaterre at kitware.com] 
Sent: Friday, May 20, 2005 11:13 AM
To: Li, George (NIH/NCI)
Cc: insight-users at itk.org
Subject: Re: GDCM image size reading


George,

	I suspect that since you are not using cmake you might not have the 
exact same flags in your project as in ITK. Thus creating conflict in 
system libraries used (singledthreaded/multithreaded & static/shared).

	As a side note cmake completely support generating VS6 workspace and

therefore there is no need to manually create them. This approach is 
bulletproof and by far strongly recommended to avoid waisting time 
carefully tuning your VS6 configuration since cmake handle that.

My 2 cents,
Mathieu

Li, George (NIH/NCI) wrote:
> Mathieu:
> 
> For DicomImageReadPrintTags I0001_1, what I did was to build 
> itk_binary_directory\Examples\IO\DicomImageReadPrintTags.dsp
> using VC++6. I'm a bit of reluctant to switch to Cmake, since
> a lot of VC GUI have been written and the application framework
> is built upon VC6. But, the different result seems bothersome 
> to me, and I don't have a good solution for this.
> 
> If you have no problem with the image file reading, I should
> exclude this possibility. BTW, I have converted the pixel data
> to my own voxel format for display, where 3D volumetric views
> are shown and the window is large enough (800x800).
> 
> Thanks,
> 
> George
> 
> 
> -----Original Message-----
> From: Mathieu Malaterre [mailto:mathieu.malaterre at kitware.com]
> Sent: Thursday, May 19, 2005 3:35 PM
> To: Li, George (NIH/NCI)
> Cc: insight-users at itk.org
> Subject: Re: GDCM image size reading
> 
> 
> George,
> 
> 	I was able to run both:
> 
> ./bin/DicomImageReadPrintTags /tmp/I0001_1
> 
> and
> 
> ./bin/DicomImageReadWrite /tmp/I0001_1 /tmp/out.dcm /tmp/out.vtk
> /tmp/rescale.dcm
> 
> I can open out.vtk with paraview without any problem and the dimension
> is 512x512.
> 
> If you are using ITK from CVS, could you do a cvs update as of today.
> Then build again. If this still does not work try rebuilding from 
> scratch (completely remove your binary directory).
> 
> Other than that I don't see where the problem could come from. Since 
> the
> image is raw you should even be able to read the file properly with 
> itkRawImageIO ...
> 
> Could you specify what tool are you using to view the DICOM file ? 
> Also
> what is the size of the window you are using to view the image ?
> 
> Thanks
> Mathieu
> 
> Li, George (NIH/NCI) wrote:
> 
>>Mathieu:
>>
>>Attached please find another tag file, and throw out the previous one. 
>>I mistyped an old button that was reserved, but really should have 
>>been removed. Sorry about the confusion.
>>
>>Let me know if this is good enough, or an image file is still needed.
>>
>>Thanks,
>>
>>George
>>
>>
>>-----Original Message-----
>>From: Mathieu Malaterre [mailto:mathieu.malaterre at kitware.com]
>>Sent: Thursday, May 19, 2005 2:48 PM
>>To: Li, George (NIH/NCI)
>>Cc: insight-users at itk.org
>>Subject: Re: GDCM image size reading
>>
>>
>>George,
>>
>>	So when your program output this:
>>
>>
>>
>>>0a22: (0028,0010) [US]  3                              Rows:    2 
>>>0a2c: (0028,0011) [US]  3                           Columns:    2 
>>
>>
>>
>>	Does this means there is no values at all ? Those values should be 
>>there (as specified by DICOM). How else are we supposed to find the 
>>dimensions of the image ?
>>
>>	By any chance could you send me one of those files ?
>>
>>Thanks
>>Mathieu
>>
>>
>>
> 
> 


More information about the Insight-users mailing list