MIT: A device for controlling the progression of Parkinson's disease

2022: Device Announcement

In late September 2022, researchers at the Massachusetts Institute of Technology and other institutes demonstrated a home device that can track a patient's movements and gait speed, which can be used to assess the severity of Parkinson's disease, disease progression and patient response to drugs.

The router-sized device Wi-Fi collects data passively using radio signals reflected from the patient's body as he moves around the house. The patient does not need to wear a gadget or change his behavior. The researchers used these devices to conduct a year-long home study involving 50 people. They showed that by using algorithms machine learning to analyze the vast amount of data they passively collected, a doctor could track the progression of Parkinson's disease and the response to medication more effectively than with periodic screening at the clinic.

A device has come out to control the progression of Parkinson's disease at home

The work uses a wireless device previously developed in Katabi's lab that analyzes radio signals reflected from a person's body. It transmits signals that consume a negligible part of the router's Wi-Fi power - these ultra-low-power signals do not interfere with other wireless devices in the house. Radio signals pass through walls and other solid objects, but are reflected from the person because of the water contained in our body. This creates a "human radar" that can track the movement of a person in a room. Radio waves always travel at the same speed, so the time it takes for a signal to reflect off the device shows how a person is moving.

A machine learning classifier is built into the device, which can detect accurate radio signals reflected from the patient, even if other people are moving around the room. Advanced algorithms use this movement data to calculate gait speed - how fast a person is walking. Since the device works in the background and works all day, every day, it can collect a huge amount of data. The researchers wanted to test whether they could apply machine learning to this data to gain insight into the disease over time.

The scientists gathered 50 participants, 34 of whom had Parkinson's disease, and conducted a year-long study of home gait measurements. In the course of the study, scientists collected more than 200 thousand individual measurements, which they averaged to smooth out the variability caused by conditions that are not related to the disease. They used statistical methods to analyze the data and found that gait speed at home could be used to effectively track the progression and severity of Parkinson's disease. For example, they showed gait speed was reduced almost twice as fast in people with Parkinson's compared to those without it.

An in-depth study of these differences yielded some key findings. For example, the researchers showed that daily fluctuations in a patient's walking speed correspond to their response to the drugs -- walking speed can improve after taking the drug and then begin to decline after a few hours when the drug's effect weakens. During the study, they learned to automate processes and reduce efforts, especially for participants and the clinical team. This knowledge will prove useful in conducting research on other neurological diseases such as Alzheimer's disease, ALS and Huntington's disease.^[1]

Notes