Determination of phase space reconstruction parameters for EEG sleep signs series of patients with obstructive sleep apnea

Abstract

The syndrome of obstructive sleep apnea (SAHOS) is a disease that affects men and women, usually aged over 40 years. This respiratory disorder is due to muscle tissue relaxation more specifically the narrowing of the pharynx, which causes breathing stops up to 10 seconds blocking the passage of air between the lungs and the external environment. The gold standard for diagnosis of sleep apnea is the polysonography by signals monitoring from the airflow, abdominal movement and electroencephalography signals (EEG). The daytime sleepiness and cognitive losses are associated with SAHOS. The SAHOS is commonly studied by spectral analysis of the EEG signals. Changes in the power spectral band ratios of the delta, theta, alpha and beta are identifying when patients with SAHOS are compared to normal control. In recent years, techniques of non-linear analysis have been employed in the evaluation of the EEG signals. For nonlinear analysis of time series by the time-delay technique embedding is necessary to estimate parameters of immersion (m) and time-delay (t). For a homogeneous sample of patients with SAHOS, time series of cortical REM channels containing apnea events and control events was evaluated by techniques for calculations of these parameters for the phase space reconstruction. Time delay and embedding dimension evaluated in two intervals, before and after SAHOS termination, are suggestible different.