Sleep Stage Classification Through HRV, Complexity Measures, and Heart Rate Asymmetry Using Generalized Estimating...

Area of Science:

• Cardiology
• Sleep Medicine
• Biomedical Engineering

Background:

• Conventional heart rate variability (HRV) and complexity measures offer limited insight into sleep stage dynamics.
• Understanding heart rhythm during sleep is crucial for diagnosing sleep-related disorders.

Purpose of the Study:

• To investigate the efficacy of heart rate asymmetry (HRA) parameters in identifying and differentiating sleep stages and transitions.
• To determine if HRA provides information beyond conventional HRV and complexity measures.

Main Methods:

• Analysis of electrocardiogram (ECG) RR interval time series from 31 polysomnographic recordings.
• Application of statistical techniques and machine learning models, including the Generalized Estimating Equation model.
• Assessment of HRA's effectiveness in distinguishing sleep stages (N2, N3) and transitions (N2-R, R-N2, W-N1).

Main Results:

• HRA models demonstrated high accuracy (AUCs ranging from 0.80 to 0.89) in identifying deep sleep stages and sleep transitions.
• Models incorporating asymmetric variables showed significant statistical performance in differentiating sleep stages.
• HRA parameters provided significant, independent information not captured by HRV and complexity measures alone.

Conclusions:

• HRA parameters offer valuable, independent insights into sleep stages and transitions.
• This approach can enhance the understanding of cardiac rhythm during sleep.
• HRA shows potential for developing more precise diagnostic tools for sleep disorders, including portable devices.