Hi Maedeh

there are a bunch of steps involved in the topology correction that you will need to run (e.g. surface inflation, quasi-homeomorphic spherical mapping, etc...). If you can synthesize a T1-ish image you can just run it through recon-all and it should do the right thing. Otherwise you can look at the recon-all.cmd from one of our example subjects such as bert and see what needs to be run post-segmentation, but there are some T1-weighted assumptions in various places

cheers
Bruce

 On Mon, 5 Aug 2019, Maedeh Khalilian wrote:


        External Email - Use Caution        

Dear Bruce,
The other day I asked you about doing topology correction on my own
segmentation(I have put my first e-mail and you response in the following
so u can track easier). according to what you suggested, I have created a
synthetic T1 volume based on my own segmentation. Now I wonder what to do
exactly as the second step. Should I use "mris_fix_topology"?
I would be grateful if you could help me with that.
Best,
Maedeh
 
"Dear FreeSurfer experts,
I know that FreeSurfer does segmentation, surface reconstruction and
topology correction through its routine pipeline using "recon-all" .
However, I have some T2w brain MR images and I have done the segmentation
and extracted WM, GM, and CSF with my own algorithm. Now I need to remove
back to back (i.e. a part of topology correction) which is caused in sulci
due to the segmentation. In other word, after doing my segmentation, I
wanna
open the closed concavities in CSF that were open originally.
Now I wanna know if I can do the topology correction using FreeSurfer while
I have done the segmentation by myself? I would be grateful if you could
help me with that.
Best,
Maedeh
 
Hi Maedeh

in principle you can, although it would be a lot easier if you just had a
reasonable quality T1. If not, you will need to create all the files that
recon-all expects to exist prior to the topology correction step (wm.mgz,
norm.mgz, orig.mgz, ....). Maybe the easiest thing to do would be to use
your gray/white/csf segmentation to create a synthetic T1 volume? Set all
the wm voxels to 110, all the gray to 70 and all the csf to 20, or
something like that?

cheers
Bruce"

_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

Reply via email to