| dc.description.abstract | Sleep studies help doctors diagnose a variety of sleep-related disorders, such as insomnia and sleep apnea. Most disorders can be managed once they are correctly diagnosed. However, sleep studies usually introduce discomfort and high cost, as patients need to go to hospitals, sleep in unfamiliar beds, and wear sensors all over the body. Such high bars may discourage patients from taking sleep studies.
In this thesis, we explore alternative solutions that make sleep studies comfortable and affordable, and at the same time, deliver clinically meaningful measurements. Specifically, we propose four systems that leverage radio frequency (RF) signals to remotely and automatically monitor the user's sleep in the user's bedroom. First, we present EZ-Sleep, the first contactless sleep sensor that automatically identifies bed entries and exits, and monitors key insomnia parameters like sleep latency and sleep efficiency. Further, we propose RF-Sleep, which demonstrates a much superior sleep stages classification accuracy than state-of-art methods. Next, we introduce DeepBreath, the first wireless respiration monitoring system that can recover the breathing signals of multiple individuals even when they right next to each other. And finally, we have BodyCompass, the first RF-based system that provides accurate sleep posture monitoring overnight.
Collectively with all four systems, we turn the user's bedroom into a contactless sleep lab. We provide doctors tools to diagnose and keep track of sleep-related symptoms without any additional efforts from the patient. | |
