Dear SPM experts
I have some questions about source reconstruction of EEG.
1. In a ERP study, can I merge EEG data of all trials from all subjects into a single mat file and average all trials, then do source reconstruction for an averaged source localization? Is it reasonable? Or we can only look at the source reconstruction one subject by one subject?
2. When I get a mesh image of source reconstruction, what is the unit of the scale of colorbar? And what is the unit of the 'T' in a data cursor? (For example, when I use the function of data cursor in Matlab to point the reddest point on the mesh and then a small table, displaying positions on X, Y, Z axis and a value named 'T', shows up in right-lower corner in the window. What is this T?).
3. When I invert EEG model to mip figure, is it necessary to cover the time period before event like -100ms or -200ms, and then to choose a interested time window after mip figure showed up? Or I can just cut the time period like 50-150ms to make a mip figure and use the same time period to get a source image on MRI template or mesh?
4. When doing group statistics on EEG source reconstruction, is it absolutely needed to use FWE p<0.05 to avoid false positive result? If some areas, known as reasonable regions based in previous studies, were significance by using uncorrected p value at 0.05, but were not significant by corrected P value or FWE. Could this be a kind of false negative result?
5. What is the exactly physical meaning of the "source"? Does it mean a generator responding to whole brain activity in a specific time period, or an area with biggest amplitude in a time period, or a node in a neural circuit?
Thank you very much
with best wishes
Po-Yu
|