A team of Northwestern investigators have developed a wearable and wireless sleep monitoring device that provides an in-depth analysis of different sleep stages and may improve the detection of sleep disorders, detailed in recent study in the Proceedings of the National Academy of Sciences.
Developed by scientists led by John A. Rogers, PhD, the Louis Simpson and Kimberly Querrey Professor of Materials Science and Engineering, Biomedical Engineering and Neurological Surgery, the novel device can classify different sleep stages — awake, non-rapid eye movement and rapid eye movement (REM) — as well as capture one’s sleep patterns and identify the effects of sleep apnea from a patient’s own home.
“What we wanted to do was create a simple, wearable device that would bridge the gap between what one can do with established sleep trackers on the wrist or the finger with what’s done in a hospital-based sleep laboratory,” said Rogers, who is also a professor of Dermatology and founding director of Northwestern’s Querrey Simpson Institute for Bioelectronics.
Unlike wrist- and finger-worn sleep monitoring devices currently on the market, Rogers’ device gently adheres to the skin and can be placed at the base of the neck, which is both comfortable and non-disruptive for the patient while they sleep and also optimizes different sleep measurements.
The device can measure a range of mechanical motions in the body, from respiratory rates and sounds such as snoring to changes in the patient’s body orientation while they sleep.
“Because the device is on the core body, we can determine whether you sleep more soundly on your right side or your left side or your stomach or your back, and that’s just far beyond what you can do if the device is on the wrist,” Rogers said.
Using haptic technology, the device can also detect the onset of snoring and alert the patient with a vibration to stimulate the patient to change their body orientation and positioning.
“It can be reactive, in that sense, to one’s sleep behaviors” Rogers said.
Because the device is also placed on the patient’s neck and close enough to the carotid artery, it can also capture heart rate and core body temperature.
These polysomnography data are then sent using Bluetooth technology to a machine learning model, called LMA-SleepNet, which then delivers a personalized and comprehensive sleep score and analysis to the patient’s wireless device, such as a phone or tablet. This sleep score can then be accessed by the patient and their physician to assess different sleep stages and identify potential sleep disorders.
“What sets our device apart isn’t just where it’s worn, it’s how smart it is,” said Yayun Du, PhD, assistant professor of Electrical and Computer Engineering at Vanderbilt University and first author of the study. “We extracted over 140 key signals related to breathing, heart rhythms and movement, then used explainable AI to accurately detect sleep stages and sleep apnea. We found that breathing patterns, more than anything else, revealed whether someone was in REM sleep or just waking up — something wrist-worn devices often miss.”
Rogers’ team tested and validated the device in a clinical sleep laboratory which included 26 healthy participants and nine participants with sleep apnea who were between the ages of 19 and 75 years.
The device can be produced at low-cost and would be comparable to the cost of current sleep monitoring devices on the market, according to Rogers.
“We can do much better home sleep tracking than is possible with a conventional wearable device. We can also capture characteristics that are closely correlated with sleep that are not captured really at all from the wrist or the finger and prominent among those characteristics have to do with respiration,” Rogers said.
Andreas Tzavelis, a student in the Medical Scientist Training Program (MSTP), was a co-author of the study.
The work was supported by the Querrey Simpson Institute for Bioelectronics at Northwestern University and the Procter & Gamble Company.
