[MINC-users] Registering 3T MRI to stereotaxic space

[Claude LEPAGE]

Paul,

I also have a new improved best bestlinreg script. :-) I gave a copy
of it to Vlad too.

To some, it may sound counter intuitive, but I'm now using -nmi
(normalized mutual information) for linear registration. It is
much more robust than xcorr and rather insensitive to variations 
in intensities if N3 is not performed a priori.  A mask is no 
longer necessary for linear registration, although cropping the 
neck usually helps. 

Claude


>
> Many thanks Vladimir!
>
> I will give it a try...
>
> @Simon: I did give bestlinreg_s a try, and it definitely does a better job 
> I previously had, i.e. mritotal!
>
> Best Regards,
>
> Paul
>
>
> On Wed, 12 Dec 2012, Vladimir S. FONOV wrote:
>
> > Hello Paul,
> >
> > you can try running "standard pipeline" - also distributed with minc-toolkit 
> > :
> >
> > 1. source the environment
> >
> > source /ipl/quarantine/experimental/2012-09-18/init.sh for bash
> > or
> > source /ipl/quarantine/experimental/2012-09-18/init.csh
> >
> > 2. run the pipeline:
> >
> >
> > /ipl/quarantine/experimental/2012-09-18/pipeline/standard_pipeline.pl 
> > <subject> <visit> <t1w.mnc> --prefix <out> --disable_nonlinear --model_dir 
> > /ipl/quarantine/models/icbm152_model_09c --model 
> > mni_icbm152_t1_tal_nlin_sym_09c
> >
> > it will perfom linear registration and brain extraction using BEaST, the 
> > output files  will following:
> >
> > <out>/<subject>/<visit>/tal/tal_xfm_<subject>_<visit>_t1w.xfm - linear 
> > transformation from native space to standard space
> >
> >
> > <out>/<subject>/<visit>/tal/tal_<subject>_<visit>_t1w.mnc - normalized t1w 
> > scan in standard space space
> >
> > <out>/<subject>/<visit>/tal/tal_comp_msk_<subject>_<visit>.mnc - brain mask.
> >
> > On 12-12-11 06:59 PM, Paul GRAVEL wrote:
> >> Dear All,
> >> 
> >> I am trying to estimate the transformation parameters to co-register an
> >> MRI acquired on the 3T MR machine here at the MNI, and it seems that
> >> mritotal doesn't do such a good job (I used it before for scans acquired
> >> on the 1.5T machine without any problems). Therefore, I was wondering if
> >> anyone would have the latest procedure (and is willing to share ;-)! )
> >> to coregister an MRI from the 3T to stereotaxic space.
> >
> >
> > -- 
> > Best regards,
> >
> > Vladimir S. FONOV ~ vladimir.fonov <at> gmail.com
> > _______________________________________________
> > MINC-users at bic.mni.mcgill.ca
> > http://www.bic.mni.mcgill.ca/mailman/listinfo/minc-users
> >
> _______________________________________________
> MINC-users at bic.mni.mcgill.ca
> http://www.bic.mni.mcgill.ca/mailman/listinfo/minc-users
>