[Insight-users] About deformation field

[Luis Ibanez]

Hi Jesse,

You can read the Vector images by simply using
the itk::ImageFileReader<>. Note that you will
have to instantiate it over a Vector image type
that matches the vector type in the file.

It is totally normal that you get different deformation
fields if you use FEM or Demons. You are optimizing
different things with each method, not to mention
that any variation of the many registration parameters
may drastically change to outcome on the deformation
field.

You shouldn't expect these two deformation fields to
be the same. Notice that you are not modeling reality
with any of the two methods. At the end, they are simply
heuristics that allow you to obtain *one plausible*
solution to the pixel to pixel association problem.

Whether the underlaying properties of the deformation
field are adequate for your problem or not,... this is
something that you must verify based on what you know
about the two images being registered.

For example, in the case of the RatLung data sets
you could expect the tissue below the diaphragm
to behave as an elastic incompressible material,
while the regions of the lungs should behave as a
compressible material.  In any case, both properties
lack real sense in this 2D case, given that we can not
know how much tissue went out of plane during the
respiration movement.

Image Analysis is not Physics, it is unlikely that
you find "the" solution. Instead you will find several
solutions that may be good enough for an application.

In that context, it is very important that you define
clearly what you want to accomplish by computing a
deformable registration, and simultaneouly stablish
how much computation time are you willing to afford
in order to get that result.



   Please let us know if you have further questions.


     Thanks


        Luis,



-----------------
Jesse wrote:
> Hi,
> 
>  
> 
> Thank you Luis,  i can use cvs now, But I have some questions:
> 
>  
> 
> 1 I can get the deformation feild as a Vector image, But how to read it, 
> i.e. I want to know displacement of some feature points.
> 
>  
> 
> 2 I have tested FEM and Demons on RatLungSlice1 and RatLungSlice2, but 
> the deformation field image seems defferent when using paraview. how to 
> explain?
> 
>  
> 
> Thanks
> 
>  
> 
> Jesse
> 
> ------------------------------------------------------------------------
> Do you Yahoo!?
> New and Improved Yahoo! Mail 
> <http://us.rd.yahoo.com/mail_us/taglines/10/*http://promotions.yahoo.com/new_mail/static/efficiency.html> 
> - Send 10MB messages!
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org
> http://www.itk.org/mailman/listinfo/insight-users