A Classification method for eye movements direction during REM sleep trained on wake electro-oculographic recordings

Area of Science:

• Neuroscience
• Sleep Science
• Ophthalmology

Background:

• Rapid eye movements (REMs) are a characteristic feature of REM sleep, but their physiological function remains largely unknown.
• Current methods for analyzing REMs primarily focus on their occurrence and duration, lacking detailed directional information.

Purpose of the Study:

• To develop and validate an automated tool for comprehensive analysis of REMs during sleep.
• To characterize the directional properties of REMs in addition to their timing.
• To provide a tool for deeper investigation into the physiological origins and functional significance of REMs.

Main Methods:

• Development of an automated procedure for REM detection and ocular artifact removal.
• Integration of a classification stage to determine the main direction of each detected REM.
• Training and validation of a supervised classifier using electrooculography (EOG) data from voluntary saccades in healthy volunteers.

Main Results:

• A novel automated tool was successfully developed for detailed REMs analysis.
• The tool provides classification of REMs based on their primary direction.
• Different classification methods were evaluated and compared for performance.

Conclusions:

• The developed automated tool offers a comprehensive description of REMs, including directional characteristics.
• This advancement provides valuable data for further research into the physiological basis and functional role of REMs.
• The tool has the potential to significantly enhance our understanding of REM sleep phenomena.