[Insight-users] help for Registration

[brian avants]

i agree with mohammad's comments.

you can use v4 registration to do either the geometric or
intensity-based approach.

relevant v4 classes:

./Modules/Registration/Metricsv4/include/itkLabeledPointSetToPointSetMetricv4.h
./Modules/Registration/Metricsv4/include/itkManifoldParzenWindowsPointSetFunction.h
./Modules/Registration/Metricsv4/include/itkJensenHavrdaCharvatTsallisPointSetToPointSetMetricv4.h
./Modules/Registration/Metricsv4/include/itkEuclideanDistancePointSetToPointSetMetricv4.h
./Modules/Registration/Metricsv4/include/itkJointHistogramMutualInformationComputeJointPDFThreaderBase.h
./Modules/Registration/Metricsv4/include/itkPointSetToPointSetMetricv4.h
./Modules/Registration/Metricsv4/include/itkExpectationBasedPointSetToPointSetMetricv4.h
./Modules/Registration/Metricsv4/include/itkPointSetFunction.h

i would recommend starting with:

tkEuclideanDistancePointSetToPointSetMetricv4.h

this test may help guide your work:

./Modules/Registration/Metricsv4/test/itkEuclideanDistancePointSetMetricRegistrationTest.cxx

as well as the more basic tests.

brian




On Thu, Sep 13, 2012 at 2:02 PM, Mohammad Mayya <mhd.mayya at gmail.com> wrote:
> Hello Massi,
>
> Can you tell what do you want the registration for? Are the two images from
> the same modality? That may help narrowing the choice.
>
> Anyway, since you reformulated your problem as geometric, you are left with
> Landmarks-based methods which are the classic example of geometric-driven
> registration. In such a setting, a number of anatomical key points
> (landmarks) are identified both in the source and the target images and a
> transformation that aims to maximize a similarity metric usually based on
> the distance between these structures is to be recovered.
> The main limitation of these methods related to the selection and extraction
> of landmarks, while their main strength is the simplicity of the
> optimization process.
> Nevertheless, when dealing with intensities (ex. images from the same
> modality), methods that seek for “visual” correspondences between the source
> and the target image can be considered. Here an explicit photometric
> correspondence of the image intensities is to be identified.
>
> Mhd
>
> On Thu, Sep 13, 2012 at 6:23 PM, Massi <Massinissa.Bandou at usherbrooke.ca>
> wrote:
>>
>> Hello!!!
>>
>> I never done any image registration before, it's my first time. I have
>> read
>> a little bit about Paul Viola & how he focused on the mutual information
>> between the source and the target image... pretty complicated.
>> Actually, I have a contour of an object (.mesh) and PET images (dicom) of
>> the same object. I want to coregistrate them and I don't know how and
>> where
>> to start.
>>
>> My first Idea was to generate a volume from the PET images, extract the
>> surface and mesh it. after, I hope to find something how to coregistrate a
>> mesh file with another mesh file!
>>
>>
>> Please any advice or hint will be welcomed!!!
>>
>> Thx
>>
>> Massi
>>
>>
>>
>> --
>> View this message in context:
>> http://itk-users.7.n7.nabble.com/help-for-Registration-tp29455.html
>> Sent from the ITK - Users mailing list archive at Nabble.com.
>> _____________________________________
>> Powered by www.kitware.com
>>
>> Visit other Kitware open-source projects at
>> http://www.kitware.com/opensource/opensource.html
>>
>> Kitware offers ITK Training Courses, for more information visit:
>> http://www.kitware.com/products/protraining.php
>>
>> 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
>
>
>
> _____________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.html
>
> Kitware offers ITK Training Courses, for more information visit:
> http://www.kitware.com/products/protraining.php
>
> 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
>