Between 7–18 million Americans suffer from sleep disordered breathing (SDB), including those who suffer from obstructive sleep apnea (OSA). Despite this high prevalence and burden of OSA, existing diagnostic techniques remain impractical for widespread screening. In this study, we introduce a new model for OSA screening and describe an at-home wearable sleep mask (named ARAM) that can robustly track the wearers’ sleep patterns. This monitoring is achieved using select sensors that enable screening and monitoring in a form-factor that can be easily self-instrumented. Based on feedback from sleep doctors and technicians, we incorporate the most valuable sensors for OSA diagnosis, while maintaining ease-of-use and comfort for the patient. We discuss the results of preliminary field trials, where both our sleep mask and a commercially available device were worn simultaneously to evaluate our device’s robustness. Based on these results, we discuss next steps for the design of the screening system, including analysis techniques that would provide more efficient screening than existing systems.
Why is this work important?
There are 7-18 million Americans suffering from sleep disorders. Among them are patients with OSA, who stop breathing either completely or partially while asleep. This is a serious condition with few reliable low-cost devices available for primary diagnosis without expert supervision.
What has been done before?
The gold standard for OSA diagnosis is overnight polysomnography (PSG). Apart from that there are many home diagnostics devices available. However, many at-home devices offer poor diagnostic quality and some of them also require expert intervention, from installation of the device to analysis of the data.
What are our contributions?
We report the construction and validation of a design for low-cost OSA screening built around a simplified screening device embedded in an at-home wearable sleep mask. This simplified screening system allows for OSA diagnosis without imposing the costs or time commitment of a full PSG.
What are the next steps?
In the next iterations of the device, we aim to improve the mechanical design and ease of use, as well as automate data analysis and screening so that the device can be evaluated in larger studies.