[Insight-users] which registration method do you recommend

[Alireza Salamy]

hello luis,
  thank you for your suggestation.I took a look at Iterativeclosestpoint 2 and it seems ok to me,but how can I Segment the fiducial from my preoperative MRI image.
and Generate with them a list of point coordinates (x,y,z)?
  the second question is that:when I do the task above,shall I make a txt file(like IterativeClosestPointFixedPoints) containing the the point coordinate ?
  thanks alot for your help
  Alireza

Luis Ibanez <luis.ibanez at kitware.com> wrote:
  

Hi Alireza,


You are dealing with at 2D / 3D registration problem.


Here is one possible way to address it:


1) Segment the fiducial from your preoperative MRI image.
Generate with them a list of point coordinates (x,y,z).
Load those coordinates into an itkPointSet.

2) Segment the black&white patters from the 2D video image
Load those coordinates into an itkPointSet.

3) Use the itkRigid3DPerspectiveTransform.h along with the
PointSetToPointSetRegistration framework that is described
in the Examples:

Insight/Examples/Patented/
IterativeClosestPoint1.cxx
IterativeClosestPoint2.cxx
IterativeClosestPoint3.cxx



Regards,


Luis


---------------------
Alireza Salamy wrote:
> *Hi all Insight users,*
> *I am working on the project which is displaying of MR images in the 
> operation room by means of Mixed reality for future use in neurosurgical 
> operation rooms.this involves fusion of optical images from a web camera 
> with preoperatively acquired MRI images.the geometrical correspondence 
> between optical and MRI images is to be carried out in two 
> steps:1-initial registration of MRI and physical reality.2-real-time 
> tracking to compensate for motion of the camera relative to the MRI 
> image.my duty deals with the first problem.*
> *I construct a phantom that is visible both in MRI images and optical 
> images.this is used together with a bitting fixation device to achieve a 
> reproducible position relative to the skull.It contain at least 3 points 
> identifiable in MRI images,e.g.containing a lipid with characteristic 
> MRI signal,as well as a structure easily recognized in the optical 
> image,e.g a simple geometrical pattern in Black and White.*
> *It would be great if you recommend me an algorithm for registration 
> between the phantom that I made and physical reality,finding the 
> geometrical relationship(transformation matrix) between the volunteer 's 
> head and the phantom.this involves identifying the MRI-visible points in 
> the phantom and using the known geometric relationship between these and 
> the optical pattern.*
> *thank you very much for your help.*
> *Alireza*
> ** 
> 
> 
> 
> 
> 
> 
> 
> ------------------------------------------------------------------------
> Food fight? 
> 
> Enjoy some healthy debate
> in the Yahoo! Answers Food & Drink Q&A. 
> 
> 
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org
> http://www.itk.org/mailman/listinfo/insight-users


 
---------------------------------
Don't pick lemons.
See all the new 2007 cars at Yahoo! Autos.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://public.kitware.com/pipermail/insight-users/attachments/20070328/4f8fc5c3/attachment.htm