AI tool predicts Parkinson’s from breathing patterns

Study: Researchers develop a diagnosis tool worn while sleeping at home.

August 25, 2022

2 Min Read

Study: Researchers develop a diagnosis tool worn while sleeping at home.

An AI tool developed by neurological experts could detect the onset of Parkinson’s disease by evaluating nocturnal breathing patterns, according to a paper in Nature Medicine.

The scientists from MIT, University of Rochester Medical Center, Mayo Clinic and Boston University developed a diagnostic tool capable of accurately predicting Parkinson’s disease based on one night of breathing data.

A passive monitoring system or a belt worn around the abdomen uses a low-power radio signal to analyze the breathing patterns of sleeping patients. The algorithm used to analyze the findings was trained on breathing data from 12,000 nights of sleep and 120,000 hours of breathing from 757 Parkinson’s disease patients.  

Twelve patients who didn’t have Parkinson’s, but later developed the disease, were flagged by the tool. The researchers are working to develop a new study to validate their results.

“All the indications so far are positive and we hope that we can start detecting Parkinson’s much earlier,” Dina Katabi, Ph.D., a computer scientist and principal investigator from MIT, told STAT News.

The link between breathing changes and Parkinson’s disease was first suggested by James Parkinson himself in the early 19th century. Currently, there are no reliable biomarkers for detecting or tracking Parkinson’s disease. The charity Parkinson’s UK suggests it’s the fastest-growing neurological disease, with some seven million people suffering globally, according to Radboud University.

The AI tool monitored the breathing patterns, blood pulses and muscle twitching during the continuous inhale and exhale phases. Since the device can be used at a patient’s home instead of strictly in a clinical setting, specialists can diagnose a much greater number of people.

Furthermore, researchers were able to distinguish between Parkinson’s and Alzheimer’s disease. With early detection, patients can begin clinical trials and test whether a drug is working. Neurological diseases typically have a high failure rate in trials because it’s difficult to evaluate symptoms and monitor the effectiveness of the treatments.

“It’s hard to say whether nocturnal breathing is going to be the measure you’re going to see a change in response to treatment. It may be more useful for diagnosis,” said Ray Dorsey, the study’s co-author and Parkinson’s disease expert at the University of Rochester. “But I think if you can get objective measures of disease in the real world, this would tell you in a shorter period of time whether a drug works.”

Get the newsletter
From automation advancements to policy announcements, stay ahead of the curve with the bi-weekly AI Business newsletter.