[Insight-users] Apply Transform

[Luis Ibanez]

Hi Oscar,



               Thanks for letting us know.


Yes,
it is common to find medical images whose direction cosines
have negative components.


Yes,
image->GetDirection() will show you what the direction cosines are.



What do you mean by the

       "12 criterions of definition of the Euler angles"

are you talking about the array of Transform parameters ?



     Please let us know,


         Luis


----------------------------------
Oscar Esteban Sanz-Dranguet wrote:
> Hi Luis,
> 
> Following your recommendations, I've found that physical coordinates of 
> Y axis are negative (¿?), so it was not a bug. The X and Z axis are 
> positive. How can I detect if an axis should be accessed with negative 
> coordinates? Should I use image->GetDirection()?
> 
> The other question I've found is 'what of the 12 criterions of 
> definition of the Euler angles is using ITK?' After a quick search on 
> the software guide, I didn't find it.
> 
> For the tests, I assumed the origin on (0,0,0) because I knew that it 
> was certainly, but I've updated my code to take the origin into account 
> :). I tested using only rotations referred to one axis and testing where 
> a white point setted on the center was going.
> 
> Many thanks,
> Oscar Esteban
> 
> Luis Ibanez escribió:
> 
>> Hi Steban,
>>
>> You code seem to be ok.
>>
>> However we are missing some important information from
>> your image.
>>
>> In particular:
>>
>>
>>      A) What are the coordinates of the Origin  ?
>>           you can get this with   image->GetOrigin()
>>
>>      B) What is the orientation of the image ?
>>           you can get this with    image->GetDirection()
>>
>>
>> Your manual computation seems to assume that
>> the origin coordinates are (0,0,0) and that the
>> direction cosines are given  by an Identity matrix.
>>
>>
>> If you want to understand the rotation of the image
>> you should first experiment with Translations set
>> to Zero.  Otherwise, as you probably know, a
>> Rotation followed by a Translation will generate
>> a different apparent center of rotation.
>>
>>
>> You are also missing to include in your email the
>> values that you are passing as angles for the
>> Euler transform.
>>
>> Also, a screenshot of the output image that lead
>> you to suspect that something is wrong, will be
>> useful for this discussion. Could you please post
>> this screenshot in a public web site ?
>>
>>
>>
>>      Thanks
>>
>>
>>           Luis
>>
>>
>> -------------------------------------------------------------------
>> On Mon, Jun 29, 2009 at 1:24 PM, Oscar Esteban Sanz-Dranguet 
>> <oesteban at die.upm.es <mailto:oesteban at die.upm.es>> wrote:
>>
>>     Hi,
>>
>>     I try to apply a rigid transform to some mri images. After
>>     SetCenter, only rotation about Y axis seems to use this center.
>>     Why I don't get the correct center of rotation about X and Z axis?
>>     (code at the end)
>>
>>     Best Regards,
>>     Oscar Esteban
>>
>>
>>     <<<<<<<<<<<<<<<<< CODE >>>>>>>>>>>>>>>>>>>>>>
>>     int main( int argc, char *argv[] )
>>      {
>>      if ( argc < 8 )
>>       {
>>       std::cerr << "Missing Parameters " << std::endl;
>>       std::cerr << "Usage: " << argv[0];
>>       std::cerr << " inputImageFile outputImageFile";
>>       std::cerr << " traslationX traslationY traslationZ";
>>       std::cerr << " rotationX rotationY rotationZ";
>>       std::cerr << std::endl;
>>       return 1;
>>       }
>>
>>      typedef itk::ImageFileReader< ImageType > ImageReaderType;
>>
>>      typedef itk::Euler3DTransform< double > TransformType;
>>
>>      // Read images
>>      ImageReaderType::Pointer  imageReader  = ImageReaderType::New();
>>      imageReader->SetFileName( argv[1] );
>>      imageReader->Update();
>>      ImageType::Pointer  image =  imageReader->GetOutput();
>>      ImageType::SpacingType spacing = image->GetSpacing();
>>      ImageType::IndexType pixelCenter = { 128, 128, 58 };
>>
>>      image->SetPixel( pixelCenter, 255);
>>
>>      TransformType::Pointer      transform     = TransformType::New();
>>
>>      TransformType::InputPointType center;
>>      center[0] = 128.5 * spacing[0];
>>      center[1] = 128.5 * spacing[1];
>>      center[2] = 58.5 *  spacing[2];
>>      transform->SetCenter( center );
>>
>>      TransformType::OutputVectorType translation;
>>      translation[0] = atof( argv[3] ) * spacing[0];
>>      translation[1] = atof( argv[4] ) * spacing[1];
>>      translation[2] = atof( argv[5] ) * spacing[2];
>>
>>      transform->SetRotation( deg2rad( atof( argv[6] ) ), deg2rad(
>>     atof( argv[7] ) ), deg2rad( atof( argv[8] ) ) );
>>      transform->SetTranslation( translation );
>>
>>      transform->Print(std::cout, 3);
>>
>>      typedef itk::ResampleImageFilter< ImageType, ImageType >       
>> ResampleFilterType;
>>      ResampleFilterType::Pointer resampler = ResampleFilterType::New();
>>
>>      resampler->SetTransform( transform );
>>      resampler->SetInput( image );
>>      resampler->SetReferenceImage( image );
>>      resampler->UseReferenceImageOn( );
>>      resampler->Update();
>>
>>      typedef itk::ImageFileWriter< ImageType >  WriterType;
>>      WriterType::Pointer writer =  WriterType::New();
>>      writer->SetFileName( argv[2] );
>>      writer->SetInput( resampler->GetOutput() );
>>      writer->Update();
>>
>>      return EXIT_SUCCESS;
>>      }
>>
>>
>>     _____________________________________
>>     Powered by www.kitware.com <http://www.kitware.com>
>>
>>     Visit other Kitware open-source projects at
>>     http://www.kitware.com/opensource/opensource.html
>>
>>     Please keep messages on-topic and check the ITK FAQ at:
>>     http://www.itk.org/Wiki/ITK_FAQ
>>
>>     Follow this link to subscribe/unsubscribe:
>>     http://www.itk.org/mailman/listinfo/insight-users
>>
>>
>>
>> -- 
>> This message has been scanned for viruses and
>> dangerous content by *MailScanner* <http://www.mailscanner.info/>, and is
>> believed to be clean. 
> 
>