Difference between revisions of "BRAINSFit to simpleITK"

From NAMIC Wiki
Jump to: navigation, search
 
(37 intermediate revisions by 2 users not shown)
Line 3: Line 3:
 
We are developing registration module in Slicer version 4 (which is using ITKv4) for deformable registration of prostate MRI.
 
We are developing registration module in Slicer version 4 (which is using ITKv4) for deformable registration of prostate MRI.
  
We want to develop a SimpleITK prostate registration tool in Slicer4/ITKv4. This Slicer4/ITKv4 registration tool should be functional, accurate and fast implementation (i.e., comparable with the  the functionality we had in Slicer3/ITKv3, which has been evaluated previously [1]).
+
We want to develop a SimpleITK prostate registration tool in Slicer4/ITKv4. This Slicer4/ITKv4 registration tool should be functional, accurate and fast (i.e., comparable with the  the functionality we had in Slicer3/ITKv3, which has been evaluated previously [1]).
  
 
=Details on the registration approach=
 
=Details on the registration approach=
Line 10: Line 10:
 
Parameters we are using to call BRAINSFit in Slicer 3.6 [https://docs.google.com/spreadsheets/d/1eWAPVe2cSAEr3FL-AhaqXaBuG-760Th7SUzb2KGAwfE/edit#gid=0 can be found here].  
 
Parameters we are using to call BRAINSFit in Slicer 3.6 [https://docs.google.com/spreadsheets/d/1eWAPVe2cSAEr3FL-AhaqXaBuG-760Th7SUzb2KGAwfE/edit#gid=0 can be found here].  
  
Sample data can be found [http://www.slicer.org/publications/item/view/2147 can be found here].  
+
Sample data [http://www.slicer.org/publications/item/view/2147 can be found here].
  
 
=Current status=
 
=Current status=
  
Registration code on github: here
+
Registration code on github: [https://github.com/PeterBehringer/SimpleITKtraining/blob/master/BRAINSFit_to_sitk_MMI.py here]
  
 
'''Latest update: April 28, 2015'''
 
'''Latest update: April 28, 2015'''
  
Thins that are working:
+
==Things that are working==
 
* exhaustive search based initialization procedure implemented and produces good results (cross-correlation metric)
 
* exhaustive search based initialization procedure implemented and produces good results (cross-correlation metric)
* SimpleITK code for rigid+affine registration is done
 
  
Things that are implemented, but are not working:
+
==Things that are implemented, but are not working==
*
+
* SimpleITK code for rigid registration
 +
** the voxel sampling exception can be avoided by cropping the images sharply around the masks.
 +
{|
 +
|
 +
|[[Image:FixedVolume.png|thumb|300px|Fixed Volume as Reference]]
 +
|[[Image:AfterInit.png|thumb|300px|SimpleITK result after Initialization]]
 +
|[[Image:AfterRigid.png|thumb|300px|SimpleITK result after rigid registration]]
 +
|[[Image:BRAINSFitOut.png|thumb|300px|BRAINSFit result after rigid without using masks and --useCenterOfGeometryAlign]]
 +
|}
 +
** The rigid registration result is too much rotated and translated. Also the number of iteration steps varies with every computation. Reason might be that seed is set randomized and not set as done in BRAINSFit [https://github.com/BRAINSia/BRAINSTools/blob/af0ff12b8b751f9423061d42178268acab5f85f4/BRAINSCommonLib/BRAINSFitHelper.h#L375 here]. There is no option at the [http://www.itk.org/SimpleITKDoxygen/html/classitk_1_1simple_1_1ImageRegistrationMethod.html SimpleITK::ImageRegistrationMethod] to set a metric seed.
  
Things that are not implemented, but need to be implemented:
+
==Things that are not implemented, but need to be implemented==
*
+
* SimpleITK code for affine registration
 +
** setting affineOptimizer [https://github.com/BRAINSia/BRAINSTools/blob/19fa37dfbdee37deff4ccee412bb601f7a787bda/BRAINSCommonLib/genericRegistrationHelper.hxx#L158-L187 here]
 +
** setting RegistrationMethod [https://github.com/BRAINSia/BRAINSTools/blob/19fa37dfbdee37deff4ccee412bb601f7a787bda/BRAINSCommonLib/genericRegistrationHelper.hxx#L289-L353 here]
 +
* SimpleITK code for BSpline registration (started [https://github.com/PeterBehringer/SimpleITKtraining/blob/master/simpleITK_to_sitk_BSpline.py here])
  
 
=Unresolved issues out of our direct control=
 
=Unresolved issues out of our direct control=

Latest revision as of 02:33, 1 May 2015

Home < BRAINSFit to simpleITK

Goal

We are developing registration module in Slicer version 4 (which is using ITKv4) for deformable registration of prostate MRI.

We want to develop a SimpleITK prostate registration tool in Slicer4/ITKv4. This Slicer4/ITKv4 registration tool should be functional, accurate and fast (i.e., comparable with the the functionality we had in Slicer3/ITKv3, which has been evaluated previously [1]).

Details on the registration approach

Registration is applied to align preprocedural and intraoperational MR T2 image volumes. We are using masks the prostate for both image data sets. Registration is done using MMI metric with rigid, affine and B-spline stages applied in sequence. In Slicer3/BRAINSFit we use gradient descent for rigid/affine, and LBFGS for B-spline.

Parameters we are using to call BRAINSFit in Slicer 3.6 can be found here.

Sample data can be found here.

Current status

Registration code on github: here

Latest update: April 28, 2015

Things that are working

  • exhaustive search based initialization procedure implemented and produces good results (cross-correlation metric)

Things that are implemented, but are not working

  • SimpleITK code for rigid registration
    • the voxel sampling exception can be avoided by cropping the images sharply around the masks.
Fixed Volume as Reference
SimpleITK result after Initialization
SimpleITK result after rigid registration
BRAINSFit result after rigid without using masks and --useCenterOfGeometryAlign
    • The rigid registration result is too much rotated and translated. Also the number of iteration steps varies with every computation. Reason might be that seed is set randomized and not set as done in BRAINSFit here. There is no option at the SimpleITK::ImageRegistrationMethod to set a metric seed.

Things that are not implemented, but need to be implemented

  • SimpleITK code for affine registration
    • setting affineOptimizer here
    • setting RegistrationMethod here
  • SimpleITK code for BSpline registration (started here)

Unresolved issues out of our direct control

Standing SimpleITK issues

  • no API for consistent initialization of the metric seed

Standing ITKv4 issues

Standing Slicer4 issues

References

[1] Fedorov A, Tuncali K, Fennessy FM, Tokuda J, Hata N, et al. (2012) Image registration for targeted MRI-guided transperineal prostate biopsy. J Magn Reson Imaging 36: 987–992. Available: http://dx.doi.org/10.1002/jmri.23688.