Fig 2. An example rhythmic component reflecting a posterior alpha source.

Our decomposition method describes patterns of between-sensor phase coupling over frequencies and over trials by rhythmic components (see also Fig 1). These components are extracted from Fourier coefficients of electrophysiological recordings, which are obtained by a spectral analysis using multitapering. Components describe the systematic structure of Fourier coefficients over sensors, tapers, frequencies and trials and consist of four parameters sets. The frequency profile (A) describes the frequencies at which there is between-sensor phase coupling. The spatial amplitude map (B) describes which sensors have consistent between-sensor phase relations. The spatial phase maps (C) describe, per frequency, the between-sensor phase relations. The trial profile (D) describes how strongly a component is present in each trial by a single number. Importantly, the spatial amplitude map and the spatial phase maps describe phase coupling at the level of sensor-pairs, by maps at the level of individual sensors. The component depicted is a posterior alpha component, and is one of 15 components extracted from the representative dataset (see Results).