Automatic detection of eye movements in REM sleep using the electrooculogram

Area of Science:

• Neuroscience
• Sleep Science
• Biomedical Engineering

Background:

• Accurate quantification of eye movements during sleep is crucial for understanding sleep physiology.
• Rapid-eye-movement (REM) sleep is characterized by distinct electrooculogram (EOG) wave patterns.
• Existing methods for eye movement detection may be labor-intensive or lack precision.

Purpose of the Study:

• To develop and validate an automated method for detecting and counting eye movements from EOG signals during REM sleep.
• To establish a reliable tool for objective sleep stage analysis.
• To facilitate research in cardiorespiratory physiology by providing precise REM sleep data.

Main Methods:

• Formulation of an automated detection algorithm as a sequential decision process.
• Application of slope and amplitude threshold criteria for eye movement identification.
• Utilisation of digital filtering and smoothing signal processing techniques to enhance EOG data quality.
• Validation against visual scoring by human observers using infant EOG data.

Main Results:

• The automated method demonstrated high accuracy in counting eye movements.
• Strong correlation (coefficient > 0.9) between automated and visual counts.
• Regression analysis showed coefficients close to 1.0 (first-order) and 0 (zero-order), indicating excellent agreement.
• The method proved effective in analyzing EOG data from infants.

Conclusions:

• The developed automated EOG analysis method provides a reliable and efficient means for counting eye movements during REM sleep.
• This tool is expected to be valuable for differentiating REM sleep substates in cardiorespiratory physiology research.
• The high correlation with human scoring supports the clinical and research utility of this automated approach.