Validation of an Automatic Arousal Detection Algorithm for Whole-Night Sleep EEG Recordings

Arousals during sleep are transient accelerations of the EEG signal, considered to reflectsleep perturbations associated with poorer sleep quality. They are typically detected by visualinspection, which is time consuming, subjective, and prevents good comparability across scorers,studies and research centres. We developed a fully automatic algorithm which aims at detectingartefact and arousal events in whole-night EEG recordings, based on time-frequency analysis withadapted thresholds derived from individual data. We ran an automated detection of arousals over35 sleep EEG recordings in healthy young and older individuals and compared it against humanvisual detection from two research centres with the aim to evaluate the algorithm performance.Comparison across human scorers revealed a high variability in the number of detected arousals,which was always lower than the number detected automatically. Despite indexing more events,automatic detection showed high agreement with human detection as reflected by its correlationwith human raters and very good Cohen’s kappa values. Finally, the sex of participants and sleepstage did not influence performance, while age may impact automatic detection, depending on thehuman rater considered as gold standard. We propose our freely available algorithm as a reliable andtime-sparing alternative to visual detection of arousals.