Presenting Groundbreaking ANALYSES at WORLD SLEEP 2023
WORLD SLEEP 2023
Promoting impactful research to improve sleep quality
Sleep Number is presenting new data at World Sleep 2023 taking place October 20-25th, 2023 in Rio de Janeiro, Brazil. The Sleep Number presentation will showcase the results of a study focused on detecting sleep apnea, measuring blood pressure, providing cardiorespiratory insights, and assessing the risk of insomnia using sleep data collected from Sleep Number smart beds.
Effortless Detection of Sleep Apnea Using a Smart Bed
We developed a new algorithm for effortless detection of sleep apnea using Deep Learning to analyze ballistocardiogram (BCG) signals collected from Sleep Number smart beds. Using laboratory PSG data, collected over multiple years from 137 participants, we were able to classify 86 unseen individuals using AHI threshold of 15 with 71.4±3.1% sensitivity and a specificity of 83.3±1.0%. This technology has the potential to make it easier to monitor sleep apnea over a long period of time, which could help improve the health and well-being of many people.
Feasibility of unobtrusively estimating blood pressure during sleep using ballistocardiography-based pulse transit times
The goal of this research was to leverage exisiting ballisticardiography (BCG) and blood pressure (BP) data to test feasibility of unobtrusively estimating BP using BCG signals from the four load-cells and one electromechanical film. The data from 40 participants were utilized. The boosted estimates for instantaneous and mean Systolic BP had an average R2 of 0.63. Unobtrusively estimating mean SBP with BCG, which exists in commercial Sleep Number smart beds today, may be beneficial to detect cardiac risk which is characterized by absence of BP decrease during sleep.
Impact of timing and regularity on sleep and cardiorespiratory metrics: a large observational study
The placement of sleep in the 24-hour day, chronotype, and regularity of the sleep-wake cycle have been identified as sleep health domains. We explored the impact of chronotype and sleep/wake regularity on sleep duration, restful sleep duration, time to fall asleep (TTFA), percent restful sleep, mean heart rate (HR), and mean breathing rate (BR). Over 330 000 participants contributed > 64 million sleep sessions from their Sleep Number smart beds for these analyses. Overall, participants with regular sleep had lower HR and BR versus participants with irregular sleep and early-chronotype participants were 2.4 times more likely to have regular sleep. Regular sleep times were associated with better overall sleep. This study contributes to the literature that shows that regularity in sleep and early chronotype tends to lead to better overall sleep and possibly better health.
Prediction of insomnia risk using sleep data from a smart bed
Insomnia affects between 10 to 30% of the adult population. But many people with insomnia do not speak with their doctors about treatment. We set out to understand the course of insomnia using the Sleep Number smart bed as an effortless measure of sleep. We found that risk for insomnia could be detected on any given sleep session data from a smart bed. The accuracy of risk detection is substantially increased as the statistical model becomes personalized.