The do's and don'ts for electrophysiological connectivity analysis

In recent years it has been increasingly recognized that insight into the dynamics of interareal interactions is crucial for our understanding of normal and pathological brain function. Methodological developments and open source availability of advanced analysis tools have enabled the wider neuroscientific community to estimate a wide range of connectivity metrics from non-invasively obtained electrophysiological signals. Next to deciding on an appropriate analysis strategy, researchers are faced with the challenge to correctly interpret their findings. Volume conduction and electromagnetic field spread cause neuronal signals to be picked up by multiple channels at once, causing spurious estimates of connectivity. Comparison across experimental groups and conditions may be confounded by differences in univariate signal properties such as signal-to-noise ratio. I will illustrate some of these interpretational pitfalls and provide some recommendations that may need to be taken into account to improve the validity of the interpretation of EEG/MEG connectivity studies