I did some time frequency analysis (wavelet analysis) at source level (source calculated using dSPM) in the region of interest "cortex postcentral" of the Destrieux Atlas provided by Brainstorm.
I had pretty nice results but some reviewer of an article keep asking me which electrodes are involved in the ROI "cortex post central", I know there is no specific electrodes involved as it is calculated on the source which are estimated based on all 64 electrodes signals, but could you confirm this to me and maybe explain how the ROI based on an anatomic atlas "contains signals" of 64 different electrodes ?
I hope I am being clear to understand,
Than you in advance,
marie
If an ROI area in the brain is activated, there is a high chance that the most relevant electrodes are around that area and not all the 64 channels are involved. So, you may need to check the electrodes around that area.
As you well understand the activity for each vertex is the result of a linear combination of activity in the electrodes, this is know as the forward model, leadfield matrix or gain matrix an it is stored the head model file. Note that if you have a unconstrained source model, then there are 3 gains for each vertex location.
For the Scout, it is a bit more complicated, as it depends on how the TF was computed.
It is possible to compute before the scout time series from the individual time time series for each vertex, and then compute the Scout TF; or it is possible to compute the TF maps for each vertex in the Scout and then apply the scout function to generate the Scout TF. As you can see obtaining the contribution from the each sensor into the final Scout TF is not trivial.
Thank you so much to both of you so to be clear all electrodes signals is taking into accoutn in the calculus of the source of each vertex but the electrodes just above (like C3 for S1) have more weight in the calculus ?