EEG source localization

Support & Discussions
1

MNE version: 1.3.0
Operating system: macOS Sonoma 14.3.1

Dear forum members,

I have a question regarding EEG source localization. Can I perform it using EEG and individual structural MRI recordings without the information on EEG channel locations (no polhemus etc)?

Thanks!

2

You can; This is typically done using one of the many standard montages (many called “built in”) that are “idealized” locations of sensors. See this tutorial in the docs.

Should you is a harder question and that’s a bit more murky and likely depends on what system you’re using and how devoted you are to getting the sensors in the “correct” spot each time. If you tell us more about your setup (system, sensor net/cap, # electrodes), we can try to give you a bit more guidance. Do you have the MRI with the sensors in place?

3

Dear Peter,

Thank you for your response!

The participants do not have MRI yet, only EEG; I am thinking of sending them to the MRI so that source localization could be performed.

4

Apologies for missing this before. Source analysis without subject MRI and electrode locations is IMHO very hand-wavy. The factors that make me more confident about a given source analysis are:

  1. Cleaned data, mostly free of major artifacts
  2. High number of channels (128 or more)
  3. Sensor locations
  4. Individual head MRI

There are tools in FieldTrip for triangulating the sensors with a few photos of the participant wearing the net/cap. I haven’t gone down that road because we’re fortunate to have a hand scanner. Although a good GSoC project one day would be to do LiDAR and Photogrammetry with iPhone or other device for sensor triangulation.