[Insight-users] deformable registration direction?

[Luis Ibanez]

Hi Sli,


        This is, unfortunately, not surprising.


Although it will be desirable for registrations to be symmetric
(or reversible), this is not usually the case.

Most of the registration algorithms have inherit internal asymmetries.

(E.g. the basic registration framework uses the gradient of the moving
image, but not the gradient of the fixed image...).


Regards,


     Luis


-----------------
sli at 2pti.com wrote:
> I am trying to apply my 3d image volumes with into
> DeformableRegistration8. One very interesting thing happened. If I
> registrate vol 1 to vol 2, it works (runs about 50 iterations, for
> example), even though I could not have got a good result yet. I mean some
> mis-registration showed in the result. However, it I try to registrate vol
> 2 to vol 1, it does not run with the result I expected (it stoped after 1
> iteration and reached the convergence condition) and the output is the
> same to vol 2.
> 
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org
> http://www.itk.org/mailman/listinfo/insight-users
>