This review aims at introducing the different methods used to assess connectivity in MEG, with emphasis on methods based on phase coupling. We will discuss the advantages of studying brain connectivity starting from neural sources, as estimated from MEG, over studying brain connectivity directly from the measured signals. MEG has been often considered similar to EEG since both are related to the electromagnetic field generated by the currents flowing within neurons and in the surrounding medium, i.e., the brain volume. Indeed, while sharing the same elementary neuronal phenomena, each of the two techniques has its own strengths as summarized in Lopes da Silva (2013), and the physics laws that describe how these currents translate into magnetic field and electric potential show the differences between the two techniques (Hämäläinen et al., 1993). Among others, the lower sensitiveness of MEG to the properties of the conducting medium (Vorwerk et al., 2014; Stenroos and Nummenmaa, 2016) and the need for a reference signal in EEG (Chella et al., 2016a, 2017; Van de Steen et al., 2019) imply an important advantage of MEG over EEG in the assessment of brain functional connectivity.
