Smartwatches can analyse quality of sleep

4 September 2013

Researchers at the Fraunhofer Institute for Computer Graphics Research IGD have developed software for commercially available smartwatches that can collect data on sleep patterns and detect anomalies.

Currently, sleep research studies employ specially developed, very expensive intelligent watches. Doctors read out the data recorded by the watches just once a week in the research lab, which slows down analysis.

Smartwatches are the latest development in smartphone technology. They can do many of the things a smartphone can do: tell the time, receive text messages and emails and find out what’s going on in social networks. They also have acceleration sensors, so can be used in studying movement patterns and some models have health-related functions such as heart rate, workout and diet monitoring.

A sleep recognition algorithm in the new software helps to detect anomalies in sleep as soon as they occur. Information such as bedtimes and duration, and quality of sleep is derived from the watch’s sensor data and then analyzed. Patients can send the recorded data straight from their home to the lab via the smartwatch’s radio module.

Gerald Bieber, a scientist at Fraunhofer IGD, said, “Our algorithm detects movements and compares them against normal sleeping and waking patterns. The sensors register both micro-movements triggered by breathing or pulse and macro-movements such as twitches of the leg.”

Burnout from chronic sleep deficit

“For the doctor in charge of the patient’s care, a digital sleep diary like this is an important tool for diagnosing sleep disorders and for choosing the right therapy,” explains Bieber. “Sleep quality is an important indicator of burnout.” According to studies, it is chronic sleep deficit and not stress that is the real cause of burnout. There are many reasons for people having difficulty falling asleep, having interrupted sleep or having non-relaxing sleep: anything from the side effects of medication to too little movement during the day, or even just the wrong mattress.

In future, Bieber and his colleagues also want to detect unconsciousness in sleep. This is an issue that can affect diabetics and epileptics. Type 1 diabetes patients quite frequently fall into a state of hypoglycemia (low blood sugar) during the night, which can result in the patient entering a life-threatening diabetic coma.

The software installed in the smartwatch would trigger an alarm in such situations and notify family members or the patient’s doctor. The smartwatch researchers are currently in talks with hospitals and soon hope to obtain test data from coma patients, in other words real sample data for comparison purposes.

At present, the smartwatch with the Fraunhofer software is being used in a pilot study. In collaboration with Vital & Physio health resort and mattress manufacturer Malie, the scientists are studying the sleep behaviour of test subjects on back-friendly mattresses. The key issue they are investigating is whether the 'right' mattress can help people with sleep disorders and enable them to sleep soundly through the night. The information the study will yield on activity patterns and sleeping behaviour could be useful in fields such as combating stress and burnout. Fraunhofer IGD is responsible for technology development and modification within the study.

Saving electricity while you sleep

What is more, people suffering from sleep disorders will not be the only ones to benefit from the smartwatch app – it also offers homeowners and renters an opportunity to save on their electricity bills.

“Eleven percent of energy consumption comes from devices in stand-by mode. Because our sensitive algorithm is capable of detecting whether, for example, the watch wearer has fallen asleep in front of the TV, the smartwatch could then switch off the TV automatically via a radio signal. Modern televisions already contain the necessary equipment, but older models can also be retrofitted with special network outlets,” says Bieber.

In future, it will also be possible to switch off such diverse household objects as alarm systems, wireless internet routers, and lights using this technology.