Event-related oscillations in human action observation: the roles of action types and EEG baselines

Suppression of the mu rhythm (∼8-12 Hz, recorded over sensorimotor areas) has been reported as a potential EEG marker for the activation of the human mirror neuron system (MNS). However, the baseline plays a crucial role in EEG experiments, and non-homogeneous baselines could contribute to inconsistencies in the literature. Here, 24 neurotypical participants underwent EEG recording during rest and observation of two-minute silent videos showing a) Non-biological movements (moving balls), b) Non-goal-directed biological actions (simple hand movements), c) Goal-directed actions toward an object (complex hand movements), and d) Social interactions (involving three people). We assessed mu power (8-12 Hz) using two different baselines, both of which are common in the literature: eyes open in a resting condition and non-biological movements. With the resting baseline, frequency analysis showed mu suppression during the observation of simple hand movements and social interactions. With the non-biological movements baseline, there was no mu suppression, and even mu enhancement while observing complex actions. To determine the overlap of these data with areas implicated for the MNS by imaging studies, the minimum norm approach was employed to reconstruct cortical sources, and permutation analysis was used to calculate mu suppression in source space. Significant clusters of mu suppression emerged in areas attributed to the MNS, alongside clusters showing mu enhancement relating to top-down inhibitory control processes of social cognition, with both baselines. Larger clusters of mu suppression in key areas of MNS (e.g., inferior parietal lobule) were observed with the non-biological baseline than the resting baseline. We speculate that the counterintuitive mu enhancement in the frequency analysis of simple and social conditions with the non-biological baseline is mainly mediated through high alpha rhythm from the occipital sites. In conclusion, although the mu rhythm seems a promising marker for investigating the human MNS, baseline choice requires serious consideration.