STEPS FOR RECONSTRUCTING THE SURFACE
-------------------------------------


location of abeś subject directory (for helpful comparison)

/space/fs1/doris/mgh_misc/abe_anat/abe_doris


1. make subject directory
cd /space/fs1/moeller/monkey_subjects_dirs
mksubjdirs lupo

2. convert averaged anatomicals to COR format and then rotate

cd /space/fs1/moeller/cooked/anatomicals/051118Lupo/3danat
mri_convert -i avg_mprage0.5.mgh -o avg_mprage_fake1.0.mgh -iis 1 -ijs 1 -iks 1

setenv cor1 /space/fs1/moeller/cooked/anatomicals/051118Lupo/3danat/avg_cor
setenv cor2 /space/fs1/moeller/cooked/anatomicals/051118Lupo/3danat/avg_cor

setenv regfile /space/fs1/moeller/cooked/anatomicals/051118Lupo/3danat/register.dat

tkregister2 --targ $cor1 --mov $cor2 --reg $regfile


reslice volume according to registration file
-----------------------------------------------
setenv cor1 /space/fs1/moeller/cooked/anatomicals/051118Lupo/3danat/avg_cor

setenv cor2 /space/fs1/moeller/cooked/anatomicals/051118Lupo/3danat/rotate_reslice
setenv regfile /space/fs1/moeller/cooked/anatomicals/051118Lupo/3danat/register.dat

mri_vol2vol --in $cor1  --xfm  $regfile  --out  $cor2  --voxres-in-plane 1 1 --invxfm
copy rotated volume to lupo/mri/orig  
So: lupo/mri/orig should contain a volume which is rotated in stereotaxic
coordinates, with the right hemisphere on the left side.


3.  select control points in tkmedit.  run normalization.
setenv cor1 /space/fs1/moeller/monkey_subjects_dirs/lupo/mri/orig
setenv control /space/fs1/moeller/monkey_subjects_dirs/lupo/mri/tmp/control.dat
setenv cor2 /space/fs1/moeller/monkey_subjects_dirs/lupo/mri/T1

mri_normalize -f $control -no1d $cor1  $cor2

4.  copy T1s into the brain directory (since the skull stripping will
fail), and run mri_segment by hand:

mri_segment $SUBJECTS_DIR/$subject/mri/brain $SUBJECTS_DIR/$subject/mri/wm

5.  run csurf, open lupo as the subject, select pons and callosum (in
prefereces/expert preferencs/ wmfill) cutting
points, then select create surface.

6. Do edit segmentation to fix topological errors, make sure that temporal lobe
   strands are thick enough and included as wm, etc.  Then repeat step 5.

7.  create flat maps.


*****************************
email from Bruce Fischl:
mri_normalize -f $SUBJECTS_DIR/$subject/tmp/control.dat -no1d \
        $SUBJECTS_DIR/$subject/mri/orig \
        $SUBJECTS_DIR/$subject/mri/T1

then copy the T1s into the brain directory
 (since the skull stripping will
fail), and run mri_segment by hand:

mri_segment $SUBJECTS_DIR/$subject/mri/brain $SUBJECTS_DIR/$subject/mri/wm

then you should be able to pick the cc and pons cutting planes manually
and run the standard inflation/topology fixing.
