A new investigation of Massachusetts Institute of Technology (MIT) reveals that artificial intelligence could detect Parkinson’s disease by remotely monitoring a person’s breathing patterns.
Parkinson’s disease is a disorder of the central nervous system that affects movement and often causes tremors. It causes damage to nerve cells in the brain that leads to a drop in dopamine levels, which causes symptoms.
According to a report on the website of computer todayMIT scientists stated that, by training a neural network, it is possible to identify the disease before a person suffers from it.
Dina Katabi, lead author of the study, explained: “Some medical studies have shown that respiratory symptoms manifest years before motor symptoms, meaning that attributes of respiration may hold promise for risk assessment prior to Parkinson’s diagnosis”.
How did they develop artificial intelligence?
The experts trained a neural network with a multitude of data extracted from tests during sleep. To collect them, they analyzed almost 12,000 nights of breathing patterns, from 757 patients with Parkinson’s disease and about 7,000 healthy control subjects.
This neural network is basically a wall-mounted device that can be used to monitor patients at home. The equipment could work as an early warning system or for patients in the initial phase of the disease.
By testing artificial intelligence, according to the report of NewAtlasthe scientists were able to identify Parkinson’s patients with 86 percent accuracy from a single night of data, while with 12 nights it reaches up to 95 percent.
The most impressive thing about this AI is that it is capable of identifying a person who will suffer from Parkinson’s before they suffer from the disease or are diagnosed. The model was able to predict Parkinson’s with 75 percent accuracy.