New Neural Network Can Detect Parkinson’s Disease by Breathing Patterns
Thanks to a new neural network from MIT, it seems that Parkinson’s Disease can be detected simply by breathing patterns. This difficult-to-diagnose disease can now possibly be found much earlier, before major symptoms service, thus providing a greater degree of treatment early on.
This achievement came to be thanks to the efforts of Dina Katabi, PhD, a professor of electrical engineering and computer science and principal investigator at the MIT Jameel Clinic. Her team at MIT developed an artificial intelligence model that can detect Parkinson’s just by reading the breathing patterns of a patient.
The team created a neural network, a series of connected algorithms that mimic the way a human brain operates. In doing so, the neural network was able to not only determine if the disease was present but also gauge its severity and track its progression over time. It was trained by MIT Ph.D. student Yuzhe Yang and post-doctorate Yuan Yuan.
In a paper about the work in Nature Medicine, the team spans multiple universities and hospitals such as the Mayo Clinic, Boston Univerity College of Health and Rehabilitation, University of Rochester Medical Center, Rutgers Univerity, and Massachusetts General Hospital.
Check out more data science news here.
This came after years of investigating the potential use of cerebrospinal fluid and neuroimaging. But due to the invasiveness and costs associated with these tests, it was found to be unacceptable for general testing as they would require specialized medical centers — making such tests almost unobtainable to the general population.
In their findings, the team was able to demonstrate that the AI assessment of Parkinson’s could be done nightly at a patient’s home while they slept. What makes this possible is the devices created by the team which look just like a normal Wi-Fi router found in most homes. The device would then emit radio signals and then analyze the reflections of those signals from the surrounding area. This extracts the patient’s breathing patterns as data that the AI can read to make its determination.
In the paper, Katabi noted, “A relationship between Parkinson’s and breathing was noted as early as 1817, in the work of Dr. James Parkinson. This motivated us to consider the potential of detecting the disease from one’s breathing without looking at movements…Some medical studies have shown that respiratory symptoms manifest years before motor symptoms, meaning that breathing attributes could be promising for risk assessment prior to Parkinson’s diagnosis.”
As the fastest-growing neurological disease in the world, Parkinson’s is catching up with Alzheimer’s Disease, which is currently the most common neurological disorder. This finding could lead to the groundbreaking treatment of the disease since finding Parkinson’s early is often difficult and easy to miss.
Read more data science articles on OpenDataScience.com, including tutorials and guides from beginner to advanced levels! Subscribe to our weekly newsletter here and receive the latest news every Thursday. You can also get data science training on-demand wherever you are with our Ai+ Training platform. Subscribe to our fast-growing Medium Publication too, the ODSC Journal, and inquire about becoming a writer.
