Optimal Number of EEG Microstates in a Serotonergic Model of Psychosis

Abstrakt

Electroencephalographic (EEG) resting-state microstates analysis is frequently used multichannel EEG analysis, which considers EEG recording as a time series of quasi-stable topographies of electric potentials. Almost all studies based on resting-state recordings concluded that only 4 archetypal maps with unique topography are sufficient to describe most of the resting-state recording. Majority of these studies refer to the original work of Koenig et al., which postulated the number of microstates on the basis of a crossvalidation criterion. However, the optimal number of archetypal microstates is not usually re-tested.In this study, we conducted an EEG microstate analysis within a double-blind placebo-controlled study aimed at a serotonergic model of psychosis induced by psilocybin (predominantly 5HT-2A receptor agonist) in healthy volunteers. Microstates analysis showed medium size-effect as nerophysiological biomarker of schizophrenia. As psilocybin substantially alters the synchronization and the frequency spectrum of EEG signal in humans, we decided to verify the number of Class Mean Maps (CMM) in two different bandwidths by cross-validation criterion, originally used for microstate analysis.