[Insight-users] a question about itkImagePCAShapeModelEstimator

[yang.xu]

Hi Luis

  Thanks for your instant reply for my problem. I think maybe the resulting principal component is correct so

I just follow my plan to do liver segmentation. However I have met another problem recently. The final segmentation

result is incorrect. Generally speaking, the result is not like a liver(http://www.siat.ac.cn/~jiafucang/ssmdata/ result02.mhd). I just modified the example(Insight\Examples\Segmentation\GeodesicActiveContourShapePriorLevelSetImageFilter.cxx)

to 3D condition. The modified code is in bold. I think maybe the parameter that I input to the filter is not appropriate but I  don't quite understand the influence of these parameters  to the final result. Can you give me some advices about the values of these parameters?(Now initial radius is $6$ ,propagation and shape prior scaling are respectively set to $0.5$ and $0.02$ and the final result is not correct.)


.....

 ReciprocalFilterType::Pointer reciprocal = ReciprocalFilterType::New();
   const double propagationScaling = atof (argv[14]);
   const double shapePriorScaling = atof (argv[15]);
   geodesicActiveContour->SetPropagationScaling(propagationScaling);
   geodesicActiveContour->SetShapePriorScaling( shapePriorScaling);
   geodesicActiveContour->SetCurvatureScaling(1.0);
   geodesicActiveContour->SetAdvectionScaling(1.0);
   geodesicActiveContour->SetMaximumRMSError(0.005);
   geodesicActiveContour->SetNumberOfIterations(1600);
   geodesicActiveContour->SetNumberOfLayers(4);
........
 // CurvatureAnisotropicDiffusionImageFilter parameter
   smoothing->SetTimeStep(0.0625);
   smoothing->SetNumberOfIterations( 5);
   smoothing->SetConductanceParameter( 3.0);
.....
//Euler
 typedef itk::Euler3DTransform<double> TransformType;
 TransformType::Pointer transform = TransformType::New();
 shape->SetTransform( transform);

.............

Thanks for your attention!
Best wishes.

              Anthony Xu

2008-12-23
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20081223/d364d149/attachment.htm>