PhD Scientific Days 2024

Mental Health Sciences II.

Calculating sleep macrostructure indicators from spectral parameters of EEG

Előadó neve

Schneider, Bence

Neptun code

DWTSVD

Előadó munkahelye

Institute of Behavioural Sciences

Előadó telefonszáma

+36202759864

Előadó e-mail címe

schneider.bence@phd.semmelweis.hu

Az előadás címe

Calculating sleep macrostructure indicators from spectral parameters of EEG

Szerző(k) neve és munkahelye

Bence Schneider¹

1: Institute of Behavioural Sciences

Bemutatás módja

Szóbeli

Szekció

Mental Health Sciences II.

Language of the presentation

English

Preferred session

Mental Health Sciences

Összefoglaló szövege

Introduction
By employing a parametric model of EEG spectra that describes both neural oscillations and aperiodic background activity, we are able toextract parameters that provide meaningful information about sleep dynamics. The spectral exponent of the aperiodic component accurately reflects sleep depth and varies consistently among sleep stages, furthermore the oscillatoryparameters are also substantially different during non-REM and REM sleep.

Aims
Our aim is to approximate traditional macrostructure indicators sleep latency and sleep efficiency in a mathematically well-defined way, alternative to manual sleep staging.

Methods
In order to examine the temporal dynamics of the spectral parameters, a moving-window method was applied to the EEG signals with a window size of 5 minutes and window step of 4 seconds. For each window the power-spectral density was calculated, and then the model was fitted and the parameters extracted. After the parameter time series are obtained, certain threshold criteria can be applied in order to discriminate between states, from these classical sleep indicators can be approximated.

Results
The continuous spectral parameter extraction was applied to a database of 52 healthy subjects (27 female) in the age range of 4–22 years. the temporal evolution of the spectral parameters
reflect the depth and structure of sleep. The calculated indicators showed significant correlations with sleep latency (p<0.001, R=0.93) and sleep efficiency (p<0.001, R=0.75) obtained from manual scoring.
The method was also applied to a small dataset of 11 healthy control subjects and 11 psychophysiological insomnia patients, the traditional and the proposed sleep latency was also significantly correlated, also note that the calculated latency discriminated the groups better.

Conclusion
The spectral parameters of EEG are sensitive to sleep dynamics reflecting several known phenomena, thus are good candidates for the objective indexing of human sleep. Furthermore, they might provide a basis for the quantitative determination of sleep quality as an alternative to classical methods relying on subjective visual scoring.

Funding
Research supported by the Ministry of Culture and Innovation (TKP2021-EGA-25).

University

Semmelweis University

Supervisor

Dr. Bódizs Róbert

Publication of my abstract

I do not give consent to the publication of my abstract on the website of the congress.

Kind

Szabad

Status

elfogadva

Accepted presentation method

szóbeli

Előadás fájl jóváhagyás

nem rendelkezett róla

Előadó

7409

Start

11:15

End

11:25