Modeling Sleep Quality Depending on Objective Actigraphic Indicators Based on Machine Learning Methods.

Area of Science:

• Biomedical Engineering
• Sleep Medicine
• Machine Learning

Background:

• Global increase in sleep disorders like insomnia, linked to stress, anxiety, and depression.
• Poor sleep quality negatively impacts daily productivity, activity levels, and overall quality of life.
• Accurate sleep quality assessment is crucial for improving human health and longevity.

Purpose of the Study:

• To develop and validate a predictive model for assessing sleep quality.
• To utilize objective actigraphy data for sleep quality classification.
• To provide a tool for better management and treatment of sleep disturbances.

Main Methods:

• Collected 24-hour actigraphy data from 22 participants.
• Utilized objective actigraphy indicators: time in bed, sleep duration, and awakenings (number and duration).
• Applied various machine learning algorithms to build and evaluate the sleep quality classification model.

Main Results:

• The developed classification model demonstrated satisfactory accuracy, ranging from 80% to 86%.
• Objective actigraphy data proved effective in distinguishing sleep quality levels.
• The model's performance indicates its potential for practical application.

Conclusions:

• The proposed actigraphy-based model offers a reliable method for sleep quality assessment.
• Findings support the integration of this model in clinical settings for sleep disorder treatment.
• The study paves the way for enhanced sleep tracking systems and electronic health devices.