[MINC-users] linear and non linear in the jacobian

[Jason Lerch]

Hello Jamila,

we do that in the following way:

1) compute the non-linear transform.
2) extract the linear part of that transform (using lin_from_nlin)
3) compute the displacement for each transform (using minc_displacement)
4) subtract the two displacement volumes to give you the pure non-linear 
displacement.
5) compute the Jacobian for the linear and pure-nonlinear displacement 
volumes.

Two of those bits of code are not yet circulating widely, but I'd be 
happy to send them to you if you'd like.

Cheers,

Jason

Jamila Ahdidan wrote:
> Dear all,
>   Here is a question that has given me some troubles, and I have seen that SPM5 solves the issue, but what about minc???
>   Here it comes,
>   I am interested in accounting for the linear (12 parameters) and non linear transformations in the determinant of the jacobian in order to measure all the differences between groups. It seems like SPM5 can do this, but how can I do it with minc. I know that the determinant of the jacobian can be calculated applying "mincblob -determinant" to the grid file (accounting only for the non-linear normalization). The question is how can I calculate the determinant for both the linear and non-linear normalization?
>    
>   I hope my question is clear and that someone somewhere has an answer to it. 
>    
>   Many regards,
>   Jamila
>
>        
> ---------------------------------
> Be a better friend, newshound, and know-it-all with Yahoo! Mobile.  Try it now.
> _______________________________________________
> MINC-users at bic.mni.mcgill.ca
> http://www2.bic.mni.mcgill.ca/mailman/listinfo/minc-users
>