[Auto sleep staging and sleep quality estimation based on BP neural network]

Area of Science:

• Neuroscience
• Biomedical Engineering
• Artificial Intelligence

Background:

• Assessing sleep quality is crucial for diagnosing sleep disorders and overall health.
• Traditional sleep scoring methods can be time-consuming and subjective.
• Objective and automated methods for sleep quality estimation are needed.

Purpose of the Study:

• To develop and validate a 3-layer Backpropagation (BP) neural network for estimating sleep quality.
• To utilize electroencephalography (EEG) complexity and power spectrum data as input for the neural network.
• To compare the neural network's estimations with subjective sleep quality impressions.

Main Methods:

• A 3-layer BP neural network was designed.
• EEG complexity and sleep-multigraph power spectrum data were used as input vectors.
• All-night sleep-stage scoring was performed.
• Key sleep parameters (sleep period, shallow sleep, deep sleep, REM sleep, wake/sleep ratio) were defined for estimation.

Main Results:

• The neural network's estimated sleep condition aligned with the subjects' subjective impressions.
• Validation was performed using data from six cases of all-night sleep.
• The method demonstrated a high degree of accuracy in reflecting perceived sleep quality.

Conclusions:

• The developed 3-layer BP neural network provides a reliable and automated method for sleep quality estimation.
• This approach offers an objective alternative to subjective sleep assessments.
• The findings suggest the method's potential for clinical application in sleep medicine.