retinotopy with freesurfer

[this post is under frequent updating]

Retinotopy analysis consists two parts, one on high resolution structural images (segmentation, inflation, cut, etc), the other on functional images.

structural
Before you start, you need to put the structural images into certain directory hierarchy and set environment variable. Assume folder “structural” is where subjects’ structural images are. Under “structural”, you have folders “SUBJ1”, “SUBJ2”, “CON14” etc for every subject. You will have to put your structural images to “structural/SUBJ1/mri/orig/001.mgz”. If you have multiple images for this subject, use “002.mgz” etc. (If you original file format is not mgz, check out how to convert image formats).

structural
    |--SUBJ1
        |--mri
            |--orig
                |--001.mgz
                |--002.mgz

Then set environment variable in linux shell:
setenv SUBJECTS_DIR $PWD

1. Run command recon-all
   recon-all -all -subjid CON14
   to process structural image. This will run for a long time (30 hours). To see what steps are being done, checkout recon-all manual.
2. Cut occiput surface
   Run command
   tksurfer CON14 lh inflated
   to display the inflated left hemisphere.
   Rotate the brain until the medial surface is facing you.
   Then select points along calcarine fissure and press button “Cut line”.
   
   Select 3 points to define the cutting plane: 2 on medial side and 1 on lateral side. Choose a 4th point to specify which portion of surface to keep and press button “Cut plane”.
   
   
   Save (File > Patch > Save As) as file lh.occip.patch.3d.
   (To know how to cut full brain, check out this PDF)
3. Flatten occiput surface
   cd to the subject’s “surf” directory and run
   mris_flatten -w 0 -distances 20 7 lh.occip.patch.3d  lh.occip.patch.flat
   To visualize the patch, you can first load the subject’s inflated surface, then File > Patch > Load Patch …
   Flattening takes 1-2 hours.
4. Repeat step 2 and 3 for the right hemisphere.

functional

1. Create directory hierarchy
   Create directory “retinotopy”, then under this folder, create directories SUBJ01, SUBJ02, SUBJ03, CON14 etc. Each of these directories is for a subject. Under each subject’s directory, create folder “bold”. In “bold”, create folder “001”, “002” etc, they are folders for runs of a single subject.  Note, the name of run folders has to be 3-digit with padding 0s. Something like “01” or “1” won’t work. Under run folders, put the functional imaging data “f.nii” and paradigm file (rtopy.par) there. If your functional image is not a nii file, check out how to convert functional images to Nifti format.

   Under SUBJ01, create a file called “subjectname” with one line string “SUBJ1” — assuming “SUBJ1” is this subject’s name under “structural” folder. This file is to link this subject’s functional and structural data.

   Create file sessid under retinotopy. In this file each line is the folder’s name for each subject.

   retinotopy
       |--sessid
       |--SUBJ01
           |--subjectname
   
           |--bold
               |--001
                   |--f.nii
                   |--rtopy.par
               |--002
                   |--f.nii
                   |--rtopy.par
   

   Please refer to freesurfer’s wiki for detailed explanation. And the following diagram is very helpful:
   

   In rtopy.par file, write two lines
   stimtype eccen
direction pos
   or
   stimtype polar
direction neg

2. Stay in folder “retinotopy”
3. Create analysis
   Run the following command
   mkanalysis-sess.new -a rtopy -TR 2 -designtype retinotopy -paradigm rtopy.par -funcstem fmcsm5 -ncycles 8
Note: ncycles is the number of periods in each run of either ecc or mm. For example, if you have 10 steps going from a small circle to a big circle in ecc, and repeat this for 4 times, then ncycles=4.
   After running this program, you will find a directory “rtopy” newly created. There you find two files, analysis.cfg and analysis.info. You may want to change some values there for your own data (e.g. -nskip). If you already removed the extra images, you don’t need to specify -nskip.
4. Preprocess functional data
   Run command
   preproc-sess -sf sessid -fwhm 5
   or
   preproc-sess -s SUBJ01 -fwhm 5
   It takes about 40s for one run. (For xc only: you should run this command on scuttlebutt, not rumor)
5. Register against structural images
   Before and after registration, you may use tkregister-sess to view how well the functional images are aligned with structural images.
   tkregister-sess -sf sessid -regheader
   Green lines are the cortical surface. Hit button “compare” to see the alignment.
   To run registration, run command
   fslregister-sess -sf sessid
6. Run the analysis
   sfa-sess -a rtopy -sf sessid
   (for xc: use scuttlebutt. It takes ~2min)
7. View intermediate results
   sliceview-sess -sf sessid -a rtopy -c eccen -map h -slice mos
sliceview-sess -sf sessid -a rtopy -c polar -map h -slice mos
8. Run paint
   paint-sess -a rtopy -sf sessid
   It takes ~1min.
9. View final result
   surf-sess -sf sessid -a rtopy -retinotopy fieldsign -flat
surf-sess -sf sessid -a rtopy -retinotopy eccen -flat
surf-sess -sf sessid -a rtopy -retinotopy polar -flat
surf-sess -sf sessid -a rtopy -c polar -flat

---

---

写作助手，把中式英语变成专业英文

---

Want to receive new post notification? 有新文章通知我