Propagation of infra-slow and slow brain activities in electroencephalogram related to behavioral information processing

Abstract

The brain-wide propagation of neural excitability manifested in fast (>1 Hz) electroencephalogram (EEG) oscillations is known to play an important role in behavioral information processing. While fluctuations in EEG infra-slow and slow activity (ISA, 0.01–0.1 Hz; SA, 0.1–1 Hz) are also considered to reflect changes in neural excitability, it remains uncertain whether EEG ISA and SA propagate over the brain and whether such propagation relates to behavioral information processing. To investigate this, we analyzed a publicly available EEG dataset recorded during video game play. EEG data were separated into ISA and SA across various frequencies. The local phase gradient method was used to measure the propagation of ISA and SA. First, we observed significant brain-wide propagation in both ISA and SA. Notably, the temporal consistency of ISA propagation was stronger than that of theta and alpha oscillations, underscoring the significance of ISA propagation. Next, we hypothesized that if ISA and SA propagation relates to behavioral information processing, the propagation direction during erroneous behavior would deviate from that during normal behavior. We found from many EEG channels that the propagation direction during erroneous behavior deviated from those during random or normal behavior across all ISA and SA frequencies. Additionally, the number of channels showing this deviation was positively correlated with the behavioral error rate across participants, further supporting our hypothesis. This study suggests that the propagation of neural excitability at infra-slow and slow frequencies may be correlated with behavioral information processing.