The researchers used an artificial intelligence tool to sift through mountains of data from study participants to find patterns that identify the disease and determine severity.
“I like to compare our understanding of Parkinson’s to a street lamp in the night; we only get a glimpse of the disease when patients visit clinic. Moreover, the methods we use to track the disease over time are subjective,” says Ray Dorsey, a professor of neurology at the University of Rochester Medical Center, and a coauthor of the study in Nature Medicine.
“As a result, we have a very limited insight into how people with Parkinson’s disease affect people’s daily lives. This study shows that remote monitoring has the potential to identify individuals with Parkinson’s and create an objective measure of severity and progression. This could be a powerful tool to detect the disease early and conduct research more efficiently.”
Lead author Dina Katabi, a professor of electrical engineering and computer science at MIT, worked closely with researchers at the University of Rochester Medical Center Center for Health + Technology (CHeT), including Dorsey and Chris Tarolli, an assistant professor of neurology at the University of Rochester.
The study is one of several research projects exploring new ways to harness remote monitoring, smartphones, smart watches, and other technologies to improve care and advance research in Parkinson’s and other diseases.
Parkinson’s is the fastest-growing neurological disorder in the world, outpacing even Alzheimer’s. More than one million Americans are currently living with the disease. While there are rare genetic forms of the disease, many cases of Parkinson’s are likely caused by exposure to certain industrial chemicals and pesticides.
James Parkinson noted changes in breathing patterns when he first described the disease in the early 19th century. The new research takes inspiration from this 200-year-old observation and targets this symptom of the disease in an effort to see if nocturnal breathing rhythms, and changes to these rhythms over time, can be analyzed to create a digital biomarker of Parkinson’s. The study used a device that passively emits radio signals that capture breathing patterns, the pulsing of blood vessels, and muscle movement during sleep.
The researchers recruited 7,687 participants, including 757 with individuals Parkinson’s, and recorded 120,000 hours of sleeping. MIT researchers then analyzed the data using a neural network, a series of connected algorithms that can sort through vast qualities of data and search for patterns. The model was able to differentiate between volunteers with Parkinson’s and those without.
There are currently no effective biomarkers to diagnose Parkinson’s, particularly in the early stages, and track its progression. Often by the time motor symptoms of the disease first emerge and a diagnosis is made, a large percentage of the dopamine-producing neurons targeted by the disease have already died off. An early diagnosis could enable patients to start treatments earlier, potentially forestalling the progression of the disease.
More precise measurement of the progression of the disease—which can vary greatly from patient to patient—will also enable scientists to better measure if experimental therapies are working. Proven remote monitoring technologies will also allow researchers to recruit study participants more widely, measure the impact new therapies more quickly and, hopefully, find new effective treatments faster.
Additional researchers from the Mayo Clinic, Massachusetts General Hospital, and Boston University contributed to the study.
Source: University of Rochester
Leave a Reply