Analysis of Multichannel EEG Patterns During Human Sleep: A Novel Approach

Abstract

Classic visual sleep stage scoring is based on electroencephalogram (EEG) frequency band analysis of 30 s epochs and is commonly performed by highly trained medical sleep specialists using additional information from submental EMG and eye movements electrooculogram (EOG). In this study, we provide the proof-of-principle in 40 subjects that sleep stages can be consistently differentiated solely on the basis of spatial 3-channel EEG patterns based on root-mean-square (RMS) amplitudes. The polysomnographic 3-channel EEG data are pre-processed by RMS averaging over intervals of 30 s leading to spatial cortical activity patterns represented by 3-dimensional vectors. These patterns are visualized using multidimensional scaling (MDS), allowing a comparison of the spatial cortical activity patterns with the conventional visual sleep scoring system according to the American Academy of Sleep Medicine (AASM). Spatial cortical activity patterns based on RMS amplitudes naturally divide into different clusters that correspond to visually scored sleep stages. Furthermore, these clusters are reproducible between different subjects. Especially the cluster associated with the REM sleep stage seems to be very different from the one associated with the wake state. This study provides a proof-of-principle that it is possible to separate sleep stages solely by analyzing spatially distributed EEG RMS amplitudes reflecting cortical activity and without classical EEG feature extractions like power spectrum analysis.

Microstructure of cortical activity during sleep reflects respiratory events and state of daytime vigilance

“…As has already been shown previously, it is possible to discriminate sleep stages solely on the basis of spatially distributed cortical activitypatternsbymeansofRMS amplitudes [9]. However, the potential of this methodology goes beyond the possibility of isolated sleep stage analysis.…”

“…These single values recorded across the different recording channels during the same time bin reflect a spatial pattern of neuronal activity present at that time. The complete method has been described previously elsewhere [8,9] but will be summarized here: The normalized RMS values of each single EEG channel were calculated for each 30 s epoch (synchronized to the sleep stage analysis). The RMS amplitudes of the three recording channels correspond to a three-dimensional vector for each 30 s interval, whereas successive vectors form a trajectory in three-dimensional space.…”

“…These distances are a measure of how similar or dissimilar two points (i. e., EEG activation patterns) in n-dimensional space are. The more similar they are, the smaller their distance [8,9]. The intra-cluster distance within each sleep stage corresponds to the variance of neuronal patterns within this stage.…”

“…As we could show previously, besides classical visual sleep stage scoring, it is possible to separate sleep stages solely by analyzing spatially distributed threechannel electroencephalogram (EEG) root mean square (RMS) amplitudes [9]. This new approach allows qualitative and quantitative evaluations of sleep architecture.…”

Analysis and Visualization of Sleep Stages based on Deep Neural Networks

“…The cortical activity patters were represented by amplitude vectors calculated via a sliding window method (for the exact procedure see [14]). We could already demonstrate that this method enables to discriminate different sleep stages in human EEG recordings [16]. Furthermore, we could analyze the microstructure of cortical activity during sleep and found that it reflects respiratory events and state of daytime vigilance [17].…”