[Freesurfer] mri_watershed Error: Problem with WM_HALF_MIN too small

Dear FS experts, When I run recon-all, one of my subjects get this error in mri_watershed process: *WATERSHED** Sorting... first estimation of the COG coord: x=121 y=125 z=133 r=78 first estimation of the main basin volume: 2032411 voxel

[Freesurfer] Statistical question

Dear FS experts Does it make a difference whether I introduced the flag doss already in mris_preproc or only later when using mri_glmfit? In mris_preproc the flag doss after the fsgd file is not recognized (see below) mris_preproc --fsgd GCS_FSGD_lh_thickness.txt doss --target average --hemi lh

Re: [Freesurfer] Statistical question

You don't need to spec dods/doss when you run mris_preproc, only with mri_glmfit. The fsgd in mris_preproc is only used to get a list of subjects to preproc (and so it is consistent with when you do run mri_glmfit). doug Juergen Haenggi wrote: Dear FS experts Does it make a difference whet

Re: [Freesurfer] Bad Talairach

I know someone that uses something like what is below to avoid using tkregister: mri_convert $subject/mri/T1.mgz $subject/mri/T1.mnc mritotal -protocol icbm $subject/mri/T1.mnc $subject/mri/T1.xfm mv $subject/mri/transforms/talairach.xfm $subject/mri/transforms/oldtalairach.xfm cp $subject/mri/T

Re: [Freesurfer] Bad Talairach

That would just do the conversion, it won't fix the problem. Does this happen with all your subjects? How is the quality of your T1s (the one you sent looks a little funny). doug Allison Stevens wrote: I know someone that uses something like what is below to avoid using tkregister: mri

Re: [Freesurfer] percent signal change

No, sorry, the raw timecourse would have to be analyzed. Elif SIKOGLU wrote: Hello. I obtained the percent signal change using roisummary-sess. (divided the BOLD amplitude by the mean functional baseline offset to get the percent signal change.) Is there a way to get an error estimate (i.e.