Attention, Neural Oscillations and Entrainment: Novel Phenomenology and the Development of a Normalized Statistical Approach

Abstract

A crucial function of attention is to select relevant inputs from an environment filled with multiple, competing sources of information. Oscillations in the alpha band (8-12 Hz) are hypothesized to index one potential stratagem for such selection; specifically, they are thought to index suppression of irrelevant information processing. Recently, another mechanism of selection has been discussed: oscillatory entrainment to environmental rhythms. By syncing neural oscillations to environmental inputs, entrainment is thought to coordinate the firing probabilities of large neuronal assemblies to ensure optimal processing of pertinent information.;One outstanding question is to what extent these two mechanisms interact. By pairing high-density electroencephalographic recordings with a continuous, visuospatial selective attention task, we observe that entrainment to rhythmic environmental stimuli shapes the temporal profile of ongoing activity in the alpha band over occipital electrodes through the cross-frequency coupling. However, current techniques to measure this interaction are underdeveloped and rely on a biased statistical estimator. We therefore develop and validate a new technique to normalize this statistical bias, and we demonstrate its application to neural data. Additional applications to arbitrary directional data are also briefly discussed.;We conclude that cross-frequency coupling between entrainment to rhythmic environmental events and power in the alpha band constitutes a novel, physiologically relevant mechanism for shaping the temporal profile of visuospatial selective attention and that continued development of novel statistical tools to measure such cross-frequency interactions will be invaluable for constructing a full understanding of the relationship between neural oscillations and attention.