This article discusses the statistical analysis of recurrent indicators calculated for different stages of sleep in three groups of subjects: healthy people, with Parkinson's disease, and with sleep apnea. Statistical analysis included the calculation of recurrent indicators and parameters associated with the estimation of the spread of recurrent indicators for sleep stages compared with the average value for this stage and for the entire night recording. It has been shown that for patients with Parkinson's disease, the spread of the recurrent indicator for the wakefulness stage is extremely large, which makes further analysis difficult, but can be a good biomarker for early recognition of the disease. For apparently healthy subjects and patients with sleep apnea, sleep stages 3 and 4 differ most from the overnight averages, but have the smallest deviations from the average for this stage. Patients with sleep apnea have a higher dispersion for REM sleep compared to healthy subjects.
87.85.dd Brain-machine interfaces
$^1$\
$^2$Saratov State University named after N.G.Chernyshevsky