Changes in functional brain connectivity in the transition from wakefulness to sleep in different EEG bands

Abstract

The reorganization of brain connectivity patterns due to changes in conscious state is poorly understood. The aim of this study is to assess methods for characterizing brain connectivity changes in different sleep stages as compared to wakefulness, and hence identify possible biomarkers of the level of consciousness based on the topography of connectivity networks. Polysomnographic recordings were collected during a sleep experiment from five healthy young subjects. Functional coupling between electroencephalographic (EEG) signals was estimated in different EEG bands with Partial Directed Coherence (PDC) and Directed Coherence (DC), which provide a frequency domain description of directed causal dependencies among time series. Results indicate that in the theta (θ) and alpha (α) bands the number of significant connections increases in the transition from sleep to wakefulness. Moreover connectivity patterns elicited in sleep are dominated by short-range connections, in contrast to long range links connecting distant areas, which are elicited in wakefulness in the α and θ bands. An inversion in the direction of information flow from anterior-posterior to posterior-anterior is noticeable in the transition from sleep to wakefulness in the θ and α bands. The drop of the number of long-range posterior-frontal links in the α band may be a promising indicator for the descent into sleep and perhaps anesthesia.