Millions of people in the U.S. suffer from sleep disorders and other disorders such as Alzheimer and Parkinson’s disease that can affect sleep. Sleep disorders are not easy to diagnose and monitor because this would involve attaching electrodes and other sensors to the patient, which may also disrupt sleep.
Researchers at the Massachusetts Institute and Massachusetts General Hospital have come up with a new way of diagnosing and monitoring sleeping disorders without attaching sensors to the body. They have developed a sleep monitoring device that uses an advanced artificial intelligence algorithm to analyze radio signals produced by an individual and translate the signals into measurements for analyzing the different stages of sleep including light sleep and rapid eye movement (REM).
One of the researchers leading the study, Dina Katabi, says that their vision is to develop health sensors that disappear into the background and capture important health signals and metrics without interfering with the user’s behavior. This device will help researchers to know whether an individual is having deep sleep, which helps in memory consolidation.
Other researchers working with Katabi on this project include Matt Bianchi, chief of the Division of Sleep Medicine at MGH, and Tommi Jaakkola, a member of the Institute for Data, Systems, and Society at MIT. The paper’s authors are Mingmin Zhao and Shicao Yue MIT, both MIT graduate students. This technology has the potential to move people from monthly sleep monitoring and studies in a sleep lab to a daily and continuous sleeptech survey at home.
This is not the first time that Katabi and her team have developed a sleep monitoring device. They have also worked on a previous project for developing monitoring devices that can measure vital signals using radio sensors. The sensors feature a wireless device emitting low-power frequency signals. These signals bounce off the body and record slight movements using the reflected waves. A careful analysis of the waves can provide information on vital signs such as breathing and pulse rates.
Katabi and her team have also used the same approach to create a sensor known as WiGait for measuring walking speed using wireless signals. This device can help medical practitioners to predict cognitive decline, certain heart and pulmonary diseases, and other health problems. After developing these sensors, Katabi and her team thought that they could use a similar approach for developing a device that monitors sleep. The current procedure for monitoring sleep involves hooking patients up to special monitoring devices in a sleep lab. However, these devices don’t provide accurate information about the individuals’ sleep patterns and behavior because they may disrupt sleep.
The development of this sleep monitoring device presents a huge opportunity because sleep is not a well-understood subject and a good number of people have problems with it. For the project to be successful, Katabi and her team had to come up with a way of translating their measurements of body movements, pulse, and breathing rate into sleep stages. This is where the AI algorithm comes in. AI enabled the specialists to develop the algorithm that can be used to monitor sleep.