This at-home radio sensor could help monitor symptoms of patients with Parkinson’s

The router-like device can track treatments for the neurological disease, too.
An illustration of a person walking across a room, with radar bouncing off of them as a series of waves coming from a box.

The at-home monitoring device uses very weak radio waves to estimate how quickly a person walks. N. Fuller, SayoStudio, courtesy Yingcheng Liu

An at-home monitoring system might help overcome some of the challenges of caring for patients with Parkinson’s disease. Because those with the neurological condition are often elderly and may need assistance traveling, getting specialized care at medical centers can be difficult. But by installing radio sensors, researchers could track the progression of their symptoms, and even changes in medication, all without the patients having to leave the comfort of their homes. The results were described in the journal Science Translational Medicine on Wednesday.

Parkinson’s is a progressive neurological disorder that affects the entire body but may be best known for its impact on mobility—it often first shows up as muscle stiffness and can develop into tremors, weakness, and other barriers to movement.  Around a million Americans live with the disease, while only 60 percent of Parkinson’s patients on Medicare see a neurologist or other specialist. “We’re asking a lot from the patient community,” says James Beck, the chief scientific officer of the Parkinson’s Foundation, who was not involved with the study.

The standard assessment of the disease is highly subjective and time-consuming, says Yingcheng Liu, a PhD student in machine perception and healthcare at the Massachusetts Institute of Technology and the study’s lead author. “We can’t really ask patients to come to the clinic every day or every week,” Liu says. “This technology gives us the possibility to continuously monitor patients, and provide more objective assessments.”

The new study’s approach involves installing a picture-frame-sized sensor on the wall of 50 people’s homes: 34 with Parkinson’s and 16 without. The sensor produces radio waves, like an ultra-weak WiFi signal, that act as a human-detecting radar. The researchers used that radar to track walking speed, which they could average over hours or days to measure a patient’s mobility.

[Related: We once thought nostalgia was a disease, but it might be key to our survival]

“There are lots of different sensors” in development to track Parkinson’s, including smart watches and phone apps, Beck says. “What’s interesting about this approach is it’s entirely passive. … To be able to track almost in real time how a person moves around is really exciting.”

The researchers compared at-home mobility to common clinical measurements of Parkinson’s. After controlling for age and other factors that would affect gait, the study team found that a patient’s walking speed, as detected by the sensor, was closely correlated with an in-depth assessment for the disease: the motor examination section of a comprehensive test for Parkinson’s, which involves rating a patient’s hand movements, speech, and walking. Quick evaluations that doctors often use, like timing someone as they stand from a chair, don’t match as closely to this test.

The at-home monitor helps avoid what’s known as the “white coat effect.” “When we are doing a test in front of a doctor, we try to perform as well as possible,” Liu says. A brief in-clinic test might reflect someone’s best effort, but not their endurance—a bit like finding a car’s maximum speed, but not its mileage.

The at-home gait measurements also detected decreases in patients’ mobility well before clinical measurements could observe the reductions. Everyone’s walking speed declined over the yearlong study, but in Parkinson’s sufferers, it went twice as fast—progression which was undetectable on the comprehensive exam.

Gait is just one approach to measuring that daily function, and previous work with the device has shown that it can also capture the full silhouette of a person–Liu says that they’re curious to find out whether it also captures  stride length, hand movement, or other motor functions.

[Related: This AI can see people through walls. Here’s how.]

The at-home monitor could also improve how to treat the illness, the researchers say.. Although Parkinson’s is incurable, its symptoms can be managed with a drug, called levodopa.

“Basically, when the patient takes this medication, their movement and their cognitive function gets better,” Liu says. But the effect wears off at different speeds for different people. And the drug can produce side effects including involuntary movements. “I compare it to drinking coffee,” Liu says: Too little, and you’ll get a migraine; too much, and you get jittery and anxious. So patients and doctors need to figure out the optimal dose and timing. “The traditional way of doing that is writing in a journal every thirty minutes for two weeks,” Liu says. He says he tried journaling like this, just to understand what the evaluation looked like for a patient. “It’s really burdensome.”

With the daily gait measurements, a doctor might be able to see the effects of medication adjustments more clearly. In the study, “when we compare the patient’s walking speed with the diary, we find that they correlate,” Liu says. The study was observational, so the researchers didn’t adjust patients’ medication intentionally, but two patients did change their medication schedule during the course of the study. Those changes were reflected in their walking speed over each day—they had fewer dips of low-mobility. “The doctor could actually look at this fluctuation curve and try to change the dosage little by little.”

Beck says that clinical studies like this are still in the proving stage.” It could be years before it’s a widely used tool, he says, but “I think this has got a head start.”