Hemispheric data in forward solution?

💬 Support & Discussions
1

Do the forward solutions computed from mne.make_forward_solution return information about which hemisphere the data lies in? One of the utility functions I’m using looks to differentiate the number of vertices per hemisphere in the following way:

n_lhverts = len(fwd['src'][0]['vertno'])
n_rhverts = len(fwd['src'][1]['vertno'])

but fwd['src'][1] does not exist. So I’m wondering how else I might determine which source points belong to each hemisphere?

  • MNE version: 1.0.3
  • operating system: macOS Monterey 12.3.1
1 Like
2

Hello @jadrew43,

how did you create that forward model?

With our sample data, everything works as you would expect:

# %%
import mne

sample_dir = mne.datasets.sample.data_path()
fwd_fname = sample_dir / 'MEG' / 'sample' / 'sample_audvis-eeg-oct-6-fwd.fif'

fwd = mne.read_forward_solution(fwd_fname)

print(len(fwd['src'][0]['vertno']))
print(len(fwd['src'][1]['vertno']))
1 Like
3

Ah apologies, I didn’t realize that forward model was part of the sample data, for some reason my search in Finder does not show that file when searching fwd.fif.

I used fwd = mne.make_forward_solution(raw.info, trans=None, src=src, bem=sphere) to make the solution myself.

Reading the file instead of computing it works as expected, thank you!

4

Oh, no no no, don’t use this one unless you’re actually working with our sample dataset! :sweat_smile: if you’re using your own data, you must create the forward model from scratch.

How did you generate src?

5

I am using the sample dataset!

fwd = mne.read_forward_solution(fwd_fname)
src = fwd['src']

If I want to use data from all MEG/EEG sensors would I use sample_audvis_meg-eeg-oct-6-fwd.fif?

6

I would guess so, yes!

1 Like