Missing BEM sub folder freesurfer

Hi, I wanted to perform source localization of my result and for that I would need to work with free surfer. I performed the MRI reconstruction with Free surfer on my virtual box, the process completed in 3 hrs, however I have BEM folder missing. The terminal result shows that the process was completed without any errors, yet I am unable to see the BEM subfolder? Why is it so? The MNE tutorial says and I quote:

`The first processing stage is the creation of various surface reconstructions. Usually a full FreeSurfer reconstruction is obtained by the following commands:

my_subject=sample

my_NIfTI=/path/to/NIfTI.nii.gz

recon-all -i $my_NIfTI -s $my_subject -all

where i stands for “input” and s for “subject”. Executing this will create the folder $SUBJECTS_DIR/sample and populate it with several subfolders ( bem , label , mri , etc).`

However I don’t get a folder called BEM after the reconstruction is over? Is this a problem? Are there further steps to this that I am missing?
What should I do to fix this?


Hi, You could use MNE to generate that.
Look at mne.bem.make_watershed_bem — MNE 0.24.1 documentation

Hi Barry, Thank you for the response. Could you help me with the documentation to compute forward model with MNE in a sequential order? I went through the available tutorials and they are really confusing as to which step I need to approach when.

what is trans.fif, how do I generate it? These are all confusing.

Any sequential documentation would help

Sorry, I have no experience with source localization.
I just no you could use to generate BEM segmentation using

mne.bem.make_watershed_bem

yes just now use mne.bem.make_watershed_bem

however you should not call your subjects sample. Use the name that is
meaningful for you (name or code of your subject)

Alex

Just after the recon-all , you can run these commands on the fly to produce the BEM surfaces and more:

# for bem surface computation (or generating the full bem directory as you asked), use 
mne watershed_bem --subject=$my_subject --overwrite
# for generating a dense scalp surface, better for MEG/EEG - MRI coregistration, use
mne_make_scalp_surfaces --subject=$my_subject --overwrite

Please make sure you have installed MNE-C (Installing MNE-C — MNE 1.3.0 documentation)

Cheers.

Using MNE-C is not strictly necessary; this step:

can be done with MNE-Python; the command is mne make_scalp_surfaces.

https://mne.tools/stable/generated/commands.html#mne-make-scalp-surfaces

And you can even do that from within Python via mne.bem.make_scalp_surfaces():
https://mne.tools/stable/generated/mne.bem.make_scalp_surfaces.html

Best wishes,
Richard

2 Likes