Interdependence of electroencephalographic and electrocardiographic power bands in human sleep

The estimation of functional interdependencies within the autonomous nervous system (ANS) has gained increasing importance to quantify the relationship between neuronal and cardiac activity. The present paper introduces a biomedical signal processing approach based on computational resources to serve as a supportive tool in the characterisation of sleep. By making use of linear methods and statistical tests upon 10 electroencephalographic (EEG) and electrocardiographic (ECG) overnight sleep recordings; the computation of cross-correlation and wavelet coherence is focused on finding distinguishable features amongst the testing subjects. The results demonstrate that mid-range neuronal oscillations (e.g.θ, α and ζ) interact with cardiac power bands at different levels, especially during unrelated REM sleep stages.