search

mne-tools / mne-python

MNE: Magnetoencephalography (MEG) and Electroencephalography (EEG) in Python
https://mne.tools
BSD 3-Clause "New" or "Revised" License
2.69k stars 1.31k forks source link

MNE - Read Forward Solution from another toolbox #10016

Open thanosDelatolas opened 2 years ago

thanosDelatolas commented 2 years ago

Read forward solution

It would be nice to read the leadfield matrix (forward solution) from another toolbox such as http://duneuro.org/.

Given the fact that the leadfield matrix is a numpy array, I think that the implementation of this feature is feasible.

Additional comments

Maybe this feature wouldn't be necessary if there was a way to solve the forward problem using FEM.

Thank you,

welcome[bot] commented 2 years ago

Hello! 👋 Thanks for opening your first issue here! ❤️ We will try to get back to you soon. 🚴🏽‍♂️

drammock commented 2 years ago

crossref to forum question: https://mne.discourse.group/t/mne-read-forward-solution-from-another-toolbox/3980

larsoner commented 2 years ago

In theory we would need more than just the lead field matrix, we'd also need an accompanying info and src (source space). But this could easily be obtained (hopefully) via something like info = mne.io.read_raw(...).info. So in theory we could have an API like:

mne.read_forward_solution_duneuro(fname, info, src)

And it's up to the user to make sure that info and src actually reflect the sensor geometry and source space, respectively, used when creating the gain matrix.

It's still unclear to me if we should actually implement this, or if it should be done at the deneuro end. @agramfort thoughts?

Either way, @thanosDelatolas as a first step, could you try making a forward solution using deneuro from the MNE-Python sample dataset, and put the script here? If we really want to support this in MNE-Python, we'll need unit tests that actually read some suitable duneuro forward and check that it's similar to the 3-layer BEM solution that we already compute. I think this would be a good first step and even if for now the code for a read_forward_solution_deneuro lives in a gist it'll allow you to proceed with your analyses.

agramfort commented 2 years ago

+1 for starting with a gist based on the sample data.

thx !

mcpiastra commented 2 years ago

@thanosDelatolas happy to help in creating the proposed script!

christian-oreilly commented 2 years ago

@thanosDelatolas @mcpiastra Did you make some progress on this topic or did you drop the matter? I love MNE-Python, but I think the lack of an easy path for FEM-based source reconstruction is currently a bit limitation. Having a way to import FEM models built outside of MNE-Python would be a huge plus.

thanosDelatolas commented 2 years ago

Hello @christian-oreilly, I actually dropped the matter because I had to focus on my thesis but I agree it would be awesome!