[Insight-users] Multi Modality Image Co Registration

Tue Jul 7 18:16:18 EDT 2009

Just to be clear. I used the term "landmark" to be the zero location
set by the radiologist in an CT/MRI exam.

On Tue, Jul 7, 2009 at 5:27 PM, Luis Ibanez<luis.ibanez at kitware.com> wrote:
>
> Hi Albert,
>
>
> 1) You can monitor the registration process by using Command/Observers
>   that listen to the IterationEvents thrown by the Optimizer.
>
>
> 2) You may find useful as well the vtkRegistrationMonitor class that
>   is available at
>
>          InsightApplications/Auxiliary/vtk
>
>   it generates surface renderings of the images being registered,
>   as the registration progresses.
>
>
>
> 3) You can initialize the registration process with landmarks,
>   by using the class
>
>          itkLandmarkBasedTransformInitializer
>
>   You will have to find the locations of the landmarks in both
>   the CT and MRI images, and then pass the coordinates of the
>   landmarks to the Initializer class.
>
>
>
> 4)  You can't view anything with ITK.   :-)
>
>    ITK doesn't provide visualization capabilities.
>
>    You may want to implement your visualization code using VTK.
>    (or see item 5. below).
>
>
> 5)  If you want to verify the quality of the registration process,
>    You may want to try VV:
>
>         http://www.creatis.insa-lyon.fr/rio/vv
>
>    or Slicer:
>
>         http://www.slicer.org
>
>
>
>   Regards,
>
>
>
>      Luis
>
>
>
> ----------------
> Albert King wrote:
>>
>> Hi Bill,
>>  Thanks for the reply.Is there registration monitoring Mechanism in
>> ITK.How to set the landmarks on ct images so that corresponding landmarks
>> appear on MRI .I mean assume that I defined 4 landmarks on CT images now i
>> want to define the lanmarks on MRI corresponding to the same slice as of ct
>> .In order to check the accuracy of registration can i view the landmark list
>> by means of itk registration algorithms.Please Help
>>  Regards
>>  Albert
>>
>> ------------------------------------------------------------------------
>> *From:* Bill Lorensen <bill.lorensen at gmail.com>
>> *To:* Albert King <albert.king84 at yahoo.com>
>> *Cc:* Insight-Users <insight-users at itk.org>
>> *Sent:* Tuesday, July 7, 2009 6:20:39 PM
>> *Subject:* Re: [Insight-users] Multi Modality Image Co Registration
>>
>> No. You do not need to perform an MPR. You may have to translate if
>> they are not landmarked the same. By landmark, I mean the zero
>> position they set in the exam. For instance in a head exam, they may
>> landmark at the bridge of the nose.
>>
>> On Tue, Jul 7, 2009 at 8:35 AM, Albert King<albert.king84 at yahoo.com
>> <mailto:albert.king84 at yahoo.com>> wrote:
>>  > Hi Bill,
>>  >
>>  > Thanks for the reply, but radiologists tell that ct and mri scans are
>>  > performed at different angluations and landmarks.In that case do I have
>> to
>>  > perform a MPR on CT so as to match the landmark of MRI.
>>  >
>>  >
>>  > Thanks and Regards
>>  > Bill
>>  >
>>  > ________________________________
>>  > From: Bill Lorensen <bill.lorensen at gmail.com
>> <mailto:bill.lorensen at gmail.com>>
>>  > To: Albert King <albert.king84 at yahoo.com
>> <mailto:albert.king84 at yahoo.com>>
>>  > Cc: insight-users at itk.org <mailto:insight-users at itk.org>
>>  > Sent: Tuesday, July 7, 2009 5:48:56 PM
>>  > Subject: Re: [Insight-users] Multi Modality Image Co Registration
>>  >
>>  > Albert,
>>  >
>>  > itk performs registration in physical space. For radiology data, the
>>  > units are normally in mm's.
>>  >
>>  > If the CT and MR data were landmarked similarly, you should not have
>>  > to modify either dataset. You do not have to resample or reorient
>>  > either volume.
>>  >
>>  > That said, the picking of parameters can be tricky.
>>  >
>>  > Bill
>>  >
>>  > On Tue, Jul 7, 2009 at 8:11 AM, Albert King<albert.king84 at yahoo.com
>> <mailto:albert.king84 at yahoo.com>> wrote:
>>  >> Hi All,
>>  >>
>>  >> I am trying to perform Image Co registration (fusion ) of CT and MRI
>> dicom
>>  >> Images .Considering my registration as rigid registration (translation
>> ,
>>  >> rotation and scaling) I have fused a single slice of a CT amd MRI
>> using
>>  >> Image registration 8.cxx.I have the following Queries.
>>  >>
>>  >> CT data set is obtained from the scanner at a different angulation
>> from
>>  >> that
>>  >> of MRI (i.e . the way we acquire CT data set is different and the way
>> we
>>  >> acquire MR data set is different).Do I have to perform Multi Planr
>>  >> reformation of the CT datatset to match  the MRI angulation .
>>  >> CT data set Matrix Size is of 512 x 512  and MRI data set Matrix Size
>> is
>>  >> 256
>>  >> x 256.Do I have to rescale so as to avoid mis registration.
>>  >> Choosing the right values of different parameters for satisfactory
>>  >> registration is a tricky job.Is there any registration monitoring
>>  >> mechanism
>>  >> in ITK.
>>  >>
>>  >>
>>  >>
>>  >> Please Help.Thanks in Advance.
>>  >>
>>  >>
>>  >>
>>  >>
>>  >>
>>  >> Regards
>>  >>
>>  >>
>>  >>
>>  >> Albert
>>  >>
>>  >> _____________________________________
>>  >> Powered by 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
>>  >>
>>  >>
>>  >
>>  >
>>
>>
>> ------------------------------------------------------------------------
>>
>> _____________________________________
>> Powered by 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
>