Continuous Monitoring Device Assesses Parkinson's Severity

A team at MIT demonstrates a new wireless device that uses machine learning to analyze a patient's movements - without the need for any wearable sensors.

“For a pharmaceutical company or a biotech company trying to develop medicines for this disease, this could greatly reduce the burden and cost and speed up the development of new therapies,” says senior author Dina Katabi  .



Transcript

00:00:01 (gentle music) - [Narrator] Nearly one million people in the US are living with Parkinson's, yet about 40% of these patients are not seen by a neurologist, or Parkinson's specialist. A team of MIT researchers led by professor Dina Katabi, would like to change this situation. They have invented a wireless device, that looks like a home wifi box. It uses machine learning to analyze patient's movements

00:00:27 from the radio waves that bounce off their bodies, without the use of any wearable sensors. The device measures Parkinson's symptoms and medication response at home. Professor Katabi and her students have collaborated with Parkinson's doctors and clinicians, to deploy their device in 50 homes to monitor patients' symptoms and their response to medication. Their results show that the device is more sensitive

00:00:52 than the medical gold standard in tracking disease progression. - One big challenge in Parkinson's disease, is to pick the proper medication dose for each patient and the device can help the physician doing so, because it shows them whether the walking speed of that patient is fluctuating, oscillating throughout the day. So, here you see an example patient. You see that the walking speed is fluctuating

00:01:17 throughout the day and here you see the same patient after the doctor changed their medications. And now you can see that the walking speed is flattened and you don't see the same oscillation that you saw on the previous figure. - This study demonstrate the power of measuring health at home. Using a simple radio wave device we were able to measure how fast, or slow individuals with Parkinson's disease

00:01:44 walk in their natural environment. And because the measure is objective, sensitive and meaningful, walking speed is associated with mortality. We can use it to evaluate new treatments for Parkinson's disease, with fewer number of participants, in a shorter period of time. - [Narrator] MIT PhD, students, Yingcheng Liu and Guo Zhang worked on developing the algorithms

00:02:07 for interpreting the output of the radio device and enabling it to track Parkinson's disease. The results of the study were published in the Science Translational Medicine Journal. (gentle music)