When EEG channels positions are aligned on anatomical MRI, is there an interpolation computed on the EEG signals in the new referential (now defined by the MRI fiducial points)?
I am sure at 99.9999999% that no interpolation of the eeg signals is computed after the cap is aligned to the mri but the researchers ask me why this is not necessary and I don't have a more accurate answer as it would add even more error...
Modifying the positions of the EEG electrodes using the menu
MRI Registration > Edit (or using a process that computes an automatic fit with the scalp or projects the electrodes on the scalp) has the objective to set the positions of the electrodes to their real positions, i.e. the 3D locations where the electrodes were when the EEG data was recorded.
Therefore there is no interpolation to compute, indeed.
If the electrodes are at their correct recording sites and you move them on the scalp to new positions for which you have no EEG data recorded, you have to compute some form of interpolation from the known values. Two approaches are possible for different purposes in Brainstorm:
- using an inverse+forward model to simulate EEG data at a set of new sites (see the tutorial Simulations for more details)
- interpolating values from the neighboring sites (similarly to the process