When using a (possibly warped) template for EEG, how do people usually co-register electrodes? Is there for example a way to measure distances along the scalp surface to make sure Cz is placed properly (mid-point between nasion & inion, mid-point between the ears)? I was helping someone that has digitized electrode locations, but no additional digitized anatomical points and no subject anatomy.
For the moment, there is no solution to help with the placement of the digitized electrodes over the template.
Would you have any suggestion to improve this first registration?
Maybe we can discuss at our next meeting. I'm not sure how Brainstorm deals with electrode positions that are in an arbitrary coordinate system. It leaves a lot of freedom in matching the electrodes and anatomy, which is what I'm thinking we could improve. Something along the lines of:
- Match the electrode Cz with its corresponding measured nasion-inion midpoint location along the scalp on the template.
- Scale the template to the head points (or electrodes) keeping this Cz match fixed.
- Warp (if that was selected) the surface to the points/electrodes without Cz fixed anymore, since the warping may alter the midpoint on the scalp anyway. But we could still constrain the warping such that the midline plane computed from the "z" electrodes match the y=0 plane for example.
Guiding the initial registration with anatomical landmarks (Cz, nasion, inion...) would be a good solution.
However, we don't have these landmarks.
For the template: You could define them in on the surface of the template and use hard-coded values in the warping code.
For the electrodes, this is trickier. Not all the caps have electrodes labelled Cz, Fz, Pz or Oz. This means that you need to ask for extra input from the user. Either asking to select an electrode and click on the corresponding point on the head surface (and multiple this by the number of landmarks you need), or simply asking for the label of the closest electrode to Cz/Fz/Oz (some might not have any).
Do you need more technical information before working on it?
Right. Actually, do we know how the standard EEG locations were placed on the template? Maybe we already have a good enough Cz position there?
I'm not very familiar with the variety of EEG electrode placement. I just know how we start by measuring the Cz position. That's why I was first wondering if it was possible to measure distances on the scalp.
Unfortunately, I won't have time to work on this.
Quite randomly... mostly by me, asking for visual confirmation to the people who shared the electrodes positions. I never thought of doing it more formally, at least with the alignment of Cz, and possibly other landmarks.
If you write code to handle this better, it could be interesting to consider the rework of all the EEG templates.
I'm not very familiar with the variety of EEG electrode placement.
You can just have a look at the existing templates.
Unfortunately, I won't have time to work on this.
I don't think I will either, at least not immediately.
I added that to the infinite todo list: https://neuroimage.usc.edu/brainstorm/Next#Anatomy