Beauchamp:PrepCortSurfModels: Difference between revisions

From OpenWetWare
Jump to navigationJump to search
No edit summary
 
(22 intermediate revisions by 2 users not shown)
Line 1: Line 1:
== Creating an Anatomical BRIK ==
== First step: creating an Anatomical BRIK ==
Surfaces are created from an anatomical (T1) MRI dataset.
Surfaces are created from an anatomical (T1) MRI dataset.
Surface creation works best if two or more T1 datasets are acquired and averaged.
Surface creation works best if two or more T1 datasets are acquired and averaged. Averaging can be done in AFNI (easier) or in FreeSurfer. To average anatomies using AFNI, use the following commands. Motion correction is applied before averaging in case the subject moved between anatomies. The base anatomy (that other anatomies are aligned to) should be the anatomy that the EPIs are best aligned with (usually the T1 collected closest in time to the EPIs).
This can be accomplished with the following commands. Motion correction is applied before averaging in case the subject moved between anatomies.
  3dAllineate -base 3DSAG_T1_2.nii -source 3DSAG_T1_1.nii -prefix ${ec}anatr1_1RegTo2 -verb -warp shift_rotate -cost mi -automask -1Dfile ${ec}anatr2toanatr1
  3dAllineate -base 3DSAG_T1_2.nii -source 3DSAG_T1_1.nii -prefix ${ec}anatr1_1RegTo2 -verb -warp shift_rotate -cost mi -automask -1Dfile ${ec}anatr2toanatr1
  3dmerge -gnzmean -nscale -prefix ${ec}anatavg 3DSAG_T1_2.nii  ${ec}anatr1_1RegTo2+orig
  3dmerge -gnzmean -nscale -prefix ${ec}anatavg 3DSAG_T1_2.nii  ${ec}anatr1_1RegTo2+orig
or
  3dAllineate -base {$ec}anatr2+orig -source {$ec}anatr2+orig -prefix ${ec}anatr1_1RegTo2 -verb -warp shift_rotate -cost mi -automask -1Dfile ${ec}anatr2toanatr1
  3dmerge -gnzmean -nscale -prefix ${ec}anatavg  {$ec}anatr2+orig  ${ec}anatr1_1RegTo2+orig
If there are three anatomies, extend as follows:
  3dAllineate -base {$ec}anatr1+orig -source {$ec}anatr2+orig -prefix ${ec}anatr2_2RegTo1 -verb -warp shift_rotate -cost mi -automask -1Dfile ${ec}anatr2toanatr1
  3dAllineate -base {$ec}anatr1+orig -source {$ec}anatr3+orig -prefix ${ec}anatr3_3RegTo1 -verb -warp shift_rotate -cost mi -automask -1Dfile ${ec}anatr3toanatr1
  3dmerge -gnzmean -nscale -prefix ${ec}anatavg  {$ec}anatr1+orig  ${ec}anatr2_2RegTo1+orig ${ec}anatr3_3RegTo1+orig


If only a single T1 was acquired, then an average dataset can be faked with the command
==  Second step: FreeSurfer preparation ==
   3dcopy 3DSAG_T1.nii ${ec}anatavg
Run the script file
  /Volumes/data/scripts/@prep_dir anat_name
'''NB: To access this script (or anything else on data1) then data1 must be mounted on your Mac.''' To run this command first change into the directory where the subject's data is stored.
e.g.
  cd /Volumes/data/UT/DB/afni
Then, type the command. e.g.
   /Volumes/data/scripts/@prep_dir DBanatavg+orig.BRIK
if you have an average anatomy or
/Volumes/data/scripts/@prep_dir DBanatr1+orig.BRIK
or
/Volumes/data/scripts/@prep_dir 3dsag_t1.nii
if you do not. Be sure to include the full file name including the .BRIK or .nii suffix. Ignore warnings issued by Freesurfer about using only one anatomical scan or incorrect group permissions.


The instructions below assume that a dataset named ECanatavg+orig has already been created.
== Next: Create the surface ==
== Automation ==
[[Beauchamp:CreateCortSurfMod|Creating Cortical Surface Models]]


All of the steps below are contained in the scripts file
==Location of the created surface==
/Volumes/data9/surface/scripts/@prep_dir
Creating a cortical surface model requires a lot of disk space (~1-2 GB per subject). Because disk space was historically limited, cortical surface models were created on a different disk than the fMRI data. e.g.
 
  /Volumes/data9/surfaces
To run this command first change into the directory where a subject's anatomy (in AFNI format) is stored in a file name ECanatavg.
for the surface and
/Volumes/data1/subjID/afni
for the fMRI data.
Because disk space is now inexpensive, in 2011 this practice was discontinued and now the cortical surface model is stored with the fMRI data is the "fs" (for freesurfer) directory:
/Volumes/data1/subjID/fs
e.g.
e.g.
  cd /Volumes/data1/UT/DB/afni
/Volumes/data1/DB/fs
Then, type
For web pages describing the previous method, please see
  /Volumes/data9/surfaces/scripts/@prep_dir ec subjname
#[[Beauchamp:PrepCortSurfModelsOLD|OLD version of Preparation for Creating Cortical Surface Models]]
Where ec is the two-letter experiment code and subjname is the subject's name in the format lastname_firstname.
#[[Beauchamp:CreateCortSurfModOLD|OLD version of Creating Cortical Surface Models]]
The scripts assumes that an anatomical named ECanatavg+orig exists in the current directory.
#[[Beauchamp:UseCortSurfModOLD|OLD version of finishing and using Cortical Surface Models]]
e.g.
  /Volumes/data9/surfaces/scripts/@prep_dir DB doe_jane


This will create surfaces in the directory doe_jane


== Initial setup in /surfaces ==
==Details of the commands==
Make subject directory in /Volumes/data9/surfaces
The @recon script performs the following steps automatically; they are included here for educational purposes but do not need to be performed.
 
mkdir lastname_firstname
 
cd into directory
 
mkdir afni
 
copy high resolution scans into afni directory
 
ie.
 
3dcopy /Volumes/data1/UT/CG/afni/CGanatr1+orig ./CGanatr1


== Creating FreeSurfer Directory Structures for Individual Subjects ==
== Creating FreeSurfer Directory Structures for Individual Subjects ==
Line 63: Line 69:
   mri_convert afni/CAavganat+orig.BRIK CA/mri/001.mgz
   mri_convert afni/CAavganat+orig.BRIK CA/mri/001.mgz


if the averaging is done in AFNI (recommended method).
if the averaging is done in AFNI (recommended method). Ignore warnings issued by Freesurfer about using only one anatomical scan
 
 
Ignore warnings issued by Freesurfer about using only one anatomical scan

Latest revision as of 07:52, 6 March 2013

First step: creating an Anatomical BRIK

Surfaces are created from an anatomical (T1) MRI dataset. Surface creation works best if two or more T1 datasets are acquired and averaged. Averaging can be done in AFNI (easier) or in FreeSurfer. To average anatomies using AFNI, use the following commands. Motion correction is applied before averaging in case the subject moved between anatomies. The base anatomy (that other anatomies are aligned to) should be the anatomy that the EPIs are best aligned with (usually the T1 collected closest in time to the EPIs).

3dAllineate -base 3DSAG_T1_2.nii -source 3DSAG_T1_1.nii -prefix ${ec}anatr1_1RegTo2 -verb -warp shift_rotate -cost mi -automask -1Dfile ${ec}anatr2toanatr1
3dmerge -gnzmean -nscale -prefix ${ec}anatavg 3DSAG_T1_2.nii  ${ec}anatr1_1RegTo2+orig

or

 3dAllineate -base {$ec}anatr2+orig -source {$ec}anatr2+orig -prefix ${ec}anatr1_1RegTo2 -verb -warp shift_rotate -cost mi -automask -1Dfile ${ec}anatr2toanatr1
 3dmerge -gnzmean -nscale -prefix ${ec}anatavg  {$ec}anatr2+orig  ${ec}anatr1_1RegTo2+orig

If there are three anatomies, extend as follows:

 3dAllineate -base {$ec}anatr1+orig -source {$ec}anatr2+orig -prefix ${ec}anatr2_2RegTo1 -verb -warp shift_rotate -cost mi -automask -1Dfile ${ec}anatr2toanatr1
 3dAllineate -base {$ec}anatr1+orig -source {$ec}anatr3+orig -prefix ${ec}anatr3_3RegTo1 -verb -warp shift_rotate -cost mi -automask -1Dfile ${ec}anatr3toanatr1
 3dmerge -gnzmean -nscale -prefix ${ec}anatavg  {$ec}anatr1+orig  ${ec}anatr2_2RegTo1+orig ${ec}anatr3_3RegTo1+orig

Second step: FreeSurfer preparation

Run the script file

 /Volumes/data/scripts/@prep_dir anat_name

NB: To access this script (or anything else on data1) then data1 must be mounted on your Mac. To run this command first change into the directory where the subject's data is stored. e.g.

 cd /Volumes/data/UT/DB/afni

Then, type the command. e.g.

 /Volumes/data/scripts/@prep_dir DBanatavg+orig.BRIK

if you have an average anatomy or

/Volumes/data/scripts/@prep_dir DBanatr1+orig.BRIK

or

/Volumes/data/scripts/@prep_dir 3dsag_t1.nii

if you do not. Be sure to include the full file name including the .BRIK or .nii suffix. Ignore warnings issued by Freesurfer about using only one anatomical scan or incorrect group permissions.

Next: Create the surface

Creating Cortical Surface Models

Location of the created surface

Creating a cortical surface model requires a lot of disk space (~1-2 GB per subject). Because disk space was historically limited, cortical surface models were created on a different disk than the fMRI data. e.g.

 /Volumes/data9/surfaces

for the surface and

/Volumes/data1/subjID/afni

for the fMRI data. Because disk space is now inexpensive, in 2011 this practice was discontinued and now the cortical surface model is stored with the fMRI data is the "fs" (for freesurfer) directory:

/Volumes/data1/subjID/fs

e.g.

/Volumes/data1/DB/fs

For web pages describing the previous method, please see

  1. OLD version of Preparation for Creating Cortical Surface Models
  2. OLD version of Creating Cortical Surface Models
  3. OLD version of finishing and using Cortical Surface Models


Details of the commands

The @recon script performs the following steps automatically; they are included here for educational purposes but do not need to be performed.

Creating FreeSurfer Directory Structures for Individual Subjects

FreeSurfer requires a specific directory structure for each subject. In addition FreeSurfer requires that subjects’ directories be listed in the $SUBJECTS_DIR. Since this is an environment variable, it can be set at anytime:

set SUBJECTS_DIR = /surfaces/subject_name/ 

The program mksubjdirs constructs the directory structure automatically.

cd $SUBJECTS_DIR
mksubjdirs subjID

Importing AFNI Files into FreeSurfer

The mri_convert program is used to convert the BRIK/HEAD files into the mgz format (gzipped MGH file),:

mri_convert $SUBJECTS_DIR/afni/subjIDanatr+orig.BRIK $SUBJECTS_DIR/subjID/mri/001.mgz

This should be repeated for every anatomical scan with the number scheme 001.mgz, 002.mgz, 003.mgz, etc.

or

 mri_convert afni/CAavganat+orig.BRIK CA/mri/001.mgz

if the averaging is done in AFNI (recommended method). Ignore warnings issued by Freesurfer about using only one anatomical scan