Sleep staging is essential for evaluating sleep and its disorders. More than one billion people worldwide exhibit inadequate sleep due to modern lifestyle behavior or physical illness, but most of them are unaware of their sleep disorders 5, leading to a major challenge for healthcare systems 6. Poor sleep is associated with excessive daytime sleepiness, impaired neurocognitive function, increased risk of accidents, and cardiovascular morbidity 1, 2, 3, 4.
Sleep plays a vital role in health and well-being by maintaining health, quality of life, and productivity. This study shows the potential of audio signal analysis as a simple, convenient, and reliable MSS estimation without contact sensors. The audio-based system was validated and produced an epoch-by-epoch (standard 30-sec segments) agreement with PSG of 87% with Cohen’s kappa of 0.7.
We trained an ensemble of one-layer, feedforward neural network classifiers fed by time-series of sleep sounds to produce real-time and offline analyses. We recorded audio signals, using non-contact microphones, of 250 patients referred to a polysomnography (PSG) study in a sleep laboratory. Our working hypothesis is that the properties of sleep sounds, such as breathing and movement, within each MSS are different. Here, we present a pioneering approach for rapid eye movement (REM), non-REM, and wake staging (macro-sleep stages, MSS) estimation based on sleep sounds analysis. Moreover, the availability of sleep studies is limited, and many people with sleep disorders remain undiagnosed. Most sleep studies today incorporate contact sensors that may interfere with natural sleep and may bias results.
0 kommentar(er)
