[MINC-development] nonlinear .xfm for t1

[Claude LEPAGE]

Hi Kristi,

You can start by looking at the non-linear registration in the 
CIVET pipeline.

https://github.com/aces/CIVET/tree/master/progs

linear registration: bestlinreg.pl
non-linear registration: best1stepnlreg.pl

You have to start with linear registration, then proceed with non-linear.
Non-linear will work well if you are close to the target.

These scripts outline registration recipes with prescribed weight,
stiffness, similarity, sub-lattice, etc. It's a good starting point.

NOTE: You can also use ANTs in the minc-toolkit as an alternative approach.

Claude

>
> Hello!
>
> Please excuse my ignorance. I would like to generate a non-linear .xfm file=
>  for my subjects' native T1 using the icbm152_nlin_asym_09c. Would the foll=
> owing command suffice or am I missing options? Not sure what values I'd nee=
> d to put for options like:
> -weight
> -stiffness
> -similarity
> -sub_lattice....etc...
>
> minctracc ${subjDir}/T1.mnc ${mni_template} ${subjDir}/nonlin.xfm -nonlinea=
> r -est_translation -est_center -clobber
>
> Thank you very much for clarifying.
>
> Best,
> Kristi