[Insight-users] information loss of images

[Luis Ibanez]

Hi Kasp,

Thanks for sending the additional information
regarding your image.

 From your metaimage header, it seems that your
data is stored in 16 bits (unsigned short).

I wonder if you are trying to read it in an
image of 8bits.  ITK will allow you to do that
but by silently truncating your data to 8 bits.
That could easily look like a corrupted image.

Please make sure that the pixel type you use
for the instantiation of the Image<> class is
capable of representing the actual data of you
dicom files.

You probably want something like:


  typedef itk::Image< unsigned short, 3 > ImageType;




    Regards,


      Luis



----------------
ja sk wrote:

> Hi Luis,
>  
> A) I'm not sure what you mean by modality, but its MRI cine from siemens
>  
> B) I created the Meta Header manually and using the MetaImageImporter
> C)I did not use any ImageFileReader/Writer for creating the .mhd file, 
> so no raw files for the volumes were created.
>  
> Well i tried DicomSeriesReadImageWrite but the images are in "time 
> series", so DicomSeriesReadImageWrite will just create a .mhd and .raw 
> for a particular slice for a time instance.
> I also tried out Jolinda's MRIConvert, and it has no means of writing 
> out a volume in this case. Any suggestions, Thank You!!
>  
> here's the mhd file:
> *********************
> NDims = 3
> DimSize = 256 256 12
> ElementSpacing = 1.25 1.25 10
> Position = 0 0 0
> ElementByteOrderMSB = True
> ElementType = MET_USHORT
> ElementDataFile = LIST 2
> mri_189.dcm
> mri_210.dcm
> mri_231.dcm
> mri_252.dcm
> mri_273.dcm
> mri_294.dcm
> mri_315.dcm
> mri_336.dcm
> mri_376.dcm
> mri_397.dcm
> mri_418.dcm
> mri_439.dcm
> **************************
>  
> Regards
> Kasp
>  
>  
> */Luis Ibanez <luis.ibanez at kitware.com>/* wrote:
> 
> 
> 
>     Hi Kasp,
> 
> 
>     The answer to your question probably
>     can be found among the answers of the
>     following new questions:
> 
> 
>     A) What image modality is stored in your DICOM files ?
> 
>     B) Did you created the MetaHeader manually ?
>     or through the ImageFileWriter ?
> 
>     C) What pixel type did you used for instantiating the
>     ImageFileReader ?
> 
>     D) What pixel type did you used for instantiating the
>     ImageFileWriter ?
> 
>     E) Can you post the MetaImage header that you got ?
> 
> 
>     Probably a post of your code will be a better
>     answer to these questions.
> 
>     FYI, The simplest way of converting DICOM to MetaImage
>     (or to any other ITK supported file format) is to use
>     the application:
> 
>     Insight/Examples/IO/
>     DicomSeriesReadImageWrite.cxx
> 
> 
>     You could also use MRIConvert
>     http://lcni.uoregon.edu/~jolinda/MRIConvert
> 
> 
> 
>     Regards,
> 
> 
> 
> 
>     Luis
> 
> 
> 
>     --------------
>     ja sk wrote:
> 
>      > Hi,
>      >
>      > I created a meta header file for a volume of dicom slices, but
>     when i
>      > tried to view the volume with the image viewer found in the Insight
>      > applications, it seems like the quality has disintegrated alot (not
>      > clear at all), as compared to reading the single dicom file. Is
>     there
>      > any way to prevent this, thanks.
>      >
>      > Regards from Kasp
>      >
> 
>