A compelling new study indicates Parkinson’s disease (PD) could be diagnosed by remotely tracking a person’s breathing patterns. Led by researchers from MIT, the study presents an AI system that uses radio waves to monitor breathing while a person sleeps.
Dina Katabi, principal investigator on the new research, said the study was inspired by 200-year-old observations from James Parkinson, the first doctor to clinically catalog signs of the degenerative neurological disease.
“A relationship between Parkinson’s and breathing was noted as early as 1817, in the work of Dr. James Parkinson,” explained Katabi. “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.”
The first step was to train a neural network on a massive nocturnal breathing dataset. Nearly 12,000 nights of breathing patterns were analyzed, from 757 Parkinson’s disease patients and around 7,000 healthy control subjects.
Testing the AI model on an independent dataset it was able to diagnose Parkinson’s patients with 86% accuracy from just one night of data. On average, the study found 12 nights of consecutive tracking could reach around 95% accuracy at diagnosing Parkinson’s.
Even more interesting is the system’s potential for diagnosing Parkinson’s disease before any motor symptoms appear. The dataset studied included data from subjects before and after a Parkinson’s diagnosis. The two sleep visits were around six years apart and the AI model could predict Parkinson’s in the undiagnosed cohort with 75% accuracy from the first set of sleep data, before a patient was diagnosed with Parkinson's.
“Currently, diagnosis of PD is based on the presence of clinical motor symptoms, which are estimated to develop after 50–80% of dopaminergic neurons have already degenerated,” the researchers write in the study. “Our system shows initial evidence that it could potentially provide risk assessment before clinical motor symptoms.”
More work is of course needed to validate the system as an early diagnostic tool, but a more immediate use could be in tracking disease progression. Other data analyzed in the study showed the AI model can track a Parkinson’s patient over 12 months and correlate changes in breathing patterns with increases in disease severity.
According to Katabi, this could have uses in a variety of contexts, from improving clinical care for patients living in remote environments to helping researchers evaluate the efficacy of new drug treatments in clinical trials.
“In terms of drug development, the results can enable clinical trials with a significantly shorter duration and fewer participants, ultimately accelerating the development of new therapies,” Katabi said. “In terms of clinical care, the approach can help in the assessment of Parkinson’s patients in traditionally underserved communities, including those who live in rural areas and those with difficulty leaving home due to limited mobility or cognitive impairment.”
It’s early days, but the researchers have already developed a wall-mounted device that can be used to monitor patients in their home. Ultimately this kind of device could act as an early-warning system for people at a higher-than-average risk of developing Parkinson’s, or early-stage patients wanting close monitoring of their disease progression.
“We envision that the system could eventually be deployed in the homes of PD patients and individuals at high risk for PD (for example, those with LRRK2 gene mutation) to passively monitor their status and provide feedback to their provider,” the researchers speculated in the new study. “If the model detects severity escalation in PD patients, or conversion to PD in high-risk individuals, the clinician could follow up with the patient to confirm the results either via telehealth or a visit to the clinic.”
The new study was published in Nature Medicine.
Source: MIT
