Achieving Accurate Automatic Sleep Staging on Manually Pre-processed EEG Data Through Synchronization Feature Extraction and Graph Metrics

Abstract

Sleep staging, the process of assigning labels to epochs of sleep, depending on the stage of sleep they belong, is an arduous, time consuming and error prone process as the initial recordings are quite often polluted by noise from different sources. To properly analyze such data and extract clinical knowledge, noise components must be removed or alleviated. In this paper a pre-processing and subsequent sleep staging pipeline for the sleep analysis of electroencephalographic signals is described. Two novel methods of functional connectivity estimation (Synchronization Likelihood/SL and Relative Wavelet Entropy/RWE) are comparatively investigated for automatic sleep staging through manually pre-processed electroencephalographic recordings. A multi-step process that renders signals suitable for further analysis is initially described. Then, two methods that rely on extracting synchronization features from electroencephalographic recordings to achieve computerized sleep staging are proposed, based on bivariate features which provide a functional overview of the brain network, contrary to most proposed methods that rely on extracting univariate time and frequency features. Annotation of sleep epochs is achieved through the presented feature extraction methods by training classifiers, which are in turn able to accurately classify new epochs. Analysis of data from sleep experiments on a randomized, controlled bed-rest study, which was organized by the European Space Agency and was conducted in the “ENVIHAB” facility of the Institute of Aerospace Medicine at the German Aerospace Center (DLR) in Cologne, Germany attains high accuracy rates, over 90% based on ground truth that resulted from manual sleep staging by two experienced sleep experts. Therefore, it can be concluded that the above feature extraction methods are suitable for semi-automatic sleep staging.

Functional Re-organization of Cortical Networks of Senior Citizens After a 24-Week Traditional Dance Program

“…The Orthogonal Discrete Wavelet Transform (ODWT) was applied and the analysis was based on the relative wavelet entropy (RWE) metric. Bi-orthogonal wavelets of 5th order were selected as the family wavelet class (Frantzidis et al, 2010, 2014b; Chriskos et al, 2018). Optimal time-frequency analysis is obtained since the mother wavelet is subjected to scaling and translation.…”

Measuring Alterations of Spontaneous EEG Neural Coupling in Alzheimer's Disease and Mild Cognitive Impairment by Means of Cross-Entropy Metrics

“…AD) was assessed by means of QDA, SVM, and DT. These techniques are widely employed for data classification from EEG recordings (Spyrou et al, 2016 ; Chriskos et al, 2018 ; Ruiz-Gómez et al, 2018 ). In order to compare our results with those of our previous study (Ruiz-Gómez et al, 2018 ), the performances of the models were described by the same statistical measures: accuracy ( Acc ), sensitivity ( Se ), specificity ( Sp ), positive predictive value ( PPV ), and negative predictive value ( NPV ).…”

Reliability of Family Dogs’ Sleep Structure Scoring Based on Manual and Automated Sleep Stage Identification

“…Even the use of semi-automatic scoring, in which the sleep-wake distinctions are made automatically and provide some quality control over the procedure, would also be beneficial and would make this time-consuming process easier. In human PSG data, there are some promising results regarding computer-assisted or automated staging technologies [8][9][10] using a wide range of signal processing methods and algorithms (for a review, see [11]).…”

“…[ 13 , 14 ]. However, there is no consensus in the literature about the training process or the feature generation, and most of the studies rely on prior knowledge to compute representative features to effectively characterize EEGs [ 9 , 15 ]. These prevent the generalizability to larger datasets from various patients with different sleep patterns, and the applicability is even more limited in non-human research.…”