Objectives: The research aimed to introduce and test an automatic method for segmentation of REM sleep into three substages, respectively characterized by: enhancement of rapid eye movements; selective enhancement of slow eye movements (SEMs); reduction in the amount of eye movements. This objective was suggested by two kinds of data reported in the literature: distinction between phasic and tonic REM sleep, and remarkable presence of SEMs during REM sleep. Methods: The following signals were recorded: 19 EEG traces, submental EMG, and two EOGs (E1-A2 and E2-A2).The automatic analysis was performed on the EOG signals recorded during the REM periods that were identified by visual scoring. Themethod applied,which was derived from a previous method for the recognition of NREM microstructure, allowed identifying and characterizing events that consisted in transient amplitude increases in either a slower (0.2–0.6Hz) or a faster (1–3Hz) component of the EOG. Segmentation was obtained by means of simple queries to a database containing the features of the events. Results: The segmentation procedure made it possible to calculate nine parameters for each REM period: the duration percentages of the three substages, and the amplitude, mean frequency, and frequency variance of the two components. Conclusions: A quantitative description of the oscillating properties of REM sleep can suggest hypotheses about currently debated ssues, regarding the thalamocortical intrinsic loop active during REM sleep, the REM sleep behaviour disorder, the reduction of vulnerability during REM sleep, the mechanisms of sleep regulation, and the complex process of sleep building.
Automatic segmentation of REM sleep into three substages
VIRGILLITO, ALESSANDRA;BARCARO, UMBERTO;L. Bonfiglio;CARBONCINI, MARIA CHIARA
2014-01-01
Abstract
Objectives: The research aimed to introduce and test an automatic method for segmentation of REM sleep into three substages, respectively characterized by: enhancement of rapid eye movements; selective enhancement of slow eye movements (SEMs); reduction in the amount of eye movements. This objective was suggested by two kinds of data reported in the literature: distinction between phasic and tonic REM sleep, and remarkable presence of SEMs during REM sleep. Methods: The following signals were recorded: 19 EEG traces, submental EMG, and two EOGs (E1-A2 and E2-A2).The automatic analysis was performed on the EOG signals recorded during the REM periods that were identified by visual scoring. Themethod applied,which was derived from a previous method for the recognition of NREM microstructure, allowed identifying and characterizing events that consisted in transient amplitude increases in either a slower (0.2–0.6Hz) or a faster (1–3Hz) component of the EOG. Segmentation was obtained by means of simple queries to a database containing the features of the events. Results: The segmentation procedure made it possible to calculate nine parameters for each REM period: the duration percentages of the three substages, and the amplitude, mean frequency, and frequency variance of the two components. Conclusions: A quantitative description of the oscillating properties of REM sleep can suggest hypotheses about currently debated ssues, regarding the thalamocortical intrinsic loop active during REM sleep, the REM sleep behaviour disorder, the reduction of vulnerability during REM sleep, the mechanisms of sleep regulation, and the complex process of sleep building.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
