​Hello Freesurfer team, 

For my thesis I'm researching the differences in grey and white matter and cortical thickness between children with and without dyslexia. My supervisor advised me to manually edit the Freesurfer-generated data. I've been editing my data as follows:  

- adding control points to (larger) WM-areas that FS did not include 

- manually editing (smaller) WM-areas through recon-editing 

- correcting the pial surface through recon-editing 

After finishing and saving my edits I reran recon-all with this command: recon-all -autorecon2-cp -autorecon -subjid (subject number). When I opened my data again in Freesurfer to check my edits, I saw that there were two big problems:

- 'distortions' in the pial surface 

- sometimes large areas of WM suddenly were excluded, even when I didn't make any edits in that area 

​It makes the edited data look much worse than the original data, which is weird I suppose.  

I've included some pictures so you can see what I mean. 

- Picture (a) is the original unedited version. 

- Picture (b) is the edited version: control points and WM edits. You can see that there's a large area in the left hemisfere excluded from the original WM area

- Picture (c) is the re-edited version of (b), control points added in that large area. This time it's included but the pial surface shows a lot of irregularities. 

Me and my supervisor tried a couple of times with other data, but we still had the same results. We also tried by only adding control points and editing manually, without the pial-corrections. 

Do you have any idea of what's going wrong in the editing proces? 

Thank you very much, 

Kind regards 

Cleo Vandervost

1st master speech and language therapy 

Katholieke Universiteit Leuven