Software system for detecting sleep and awake states
Hamburg, November 2019. Several scientific studies confirm the relationship between movement, breathing, and heart rate with sleep phases. And these parameters can be determined in a more convenient way than by the PSG. The aim of this work is to use these bio vital data and the corresponding algorithmic background to carry out an initial investigation of the sleep-wake structure as a diagnostic step in a home environment.
Maksym Gaiduk presents the recent research results.
The approach presented at the German Sleep Society Comgress in Hamburg has several unique selling points: First, it uses exclusively bio-vital signals from the human body that can be measured non-invasively: respiration, heart rate, and movement. Second, an innovative set of derived parameters for sleep/wake state detection is proposed. This selection is the result of explorative data analysis. Finally, another unique feature is the application of multinomial logistic regression to sleep/wake state detection. The results already obtained have confirmed the potential of the chosen approach.