Source localization

From Waisman Brain Imaging Wiki

It is presumed that the voltages on the scalp measured by EEG equipment are the result of active volumes within the brain generating tiny electrical currents. It would be useful to be able to interpolate the locations of the active brain volumes from the EEG pattern measured at the scalp. Source localization, also known as "the inverse EEG / MEG problem", is the craft of attempting to derive these locations, or at least a simplified representation of these volumes, as represented by one or more electrical dipoles (points within the brain having a specific location and electric field strength and spatial orientation).

The derivation is complicated by the fact that each EEG pattern can be generated by more than one arrangement of dipoles, and can therefore not be assigned to a single unique solution. Thus, any source localization software package must present a derivation that is some form of approximation, best guess, and / or compromise. Since it is a guess, source localization is also referred to as "source estimation".

This inverse problem is one of the key reasons that EEG data and PET or MRI data cannot be directly compared.