The study of brain activity often requires researchers to face the trade-off between having a higher spatial resolution or a higher temporal resolution. In particular, non-invasive electrophysiological recording techniques (e.g. EEG/MEG), allow having a high temporal resolution signal that is hardly ascribable to the activity of a specific brain region. Indeed, the observable scalp electric/magnetic field is the result of several electrical dipoles generated by neural cortical activity and, to a small extent, by deeper sources. Thus, in order to disentangle distinct origins of the recorded signal, we can use source reconstruction.
In the literature, we can find several different algorithms to perform source reconstruction, most of which requires two fundamental elements: a forward model and a boundary element model. In this first seminar, we will see how source reconstruction works, what is needed to perform it and what are the advantages and the pitfalls associated with these methods.
Join at: imt.lu/aula1