Using the body’s own sounds to diagnose Alzheimer’s and Parkinson’s before the first symptoms appear

Researchers at the École de technologie supérieure (ÉTS) are developing wearable in-ear devices that analyze body sounds to detect early signs of Alzheimer’s and Parkinson’s disease before clinical symptoms appear.

The technology involves miniaturized microphones embedded in in-ear devices that capture internal sounds such as heartbeats, breathing, and swallowing. These signals are amplified by the occlusion effect, allowing for detailed analysis of subtle physiological markers associated with neurodegenerative diseases. The research team, led by experts in multimodal health monitoring, is exploring methods like machine learning and audio source separation to disentangle overlapping signals, aiming to identify early indicators such as altered breathing patterns or eye movements linked to disease progression. Current studies involve comparing physiological signals from healthy individuals and patients with neurodegenerative conditions, including Parkinson’s disease, in collaboration with Parkinson Québec and the Université de Montréal, as well as healthy controls and Alzheimer’s patients at the Douglas Research Centre.

Why It Matters

This development could revolutionize early diagnosis of neurodegenerative diseases, which currently rely on observable symptoms that appear late in disease progression. Early detection enables timely intervention, potentially slowing disease advancement and improving quality of life. The noninvasive, continuous monitoring approach also offers a personalized tool for tracking disease progression and response to treatment, representing a significant advance in neuropsychology and clinical practice.

Background

Early signs of Parkinson’s include subtle changes in speech, breathing, and muscle control, often overlooked until more severe symptoms develop. Similarly, early Alzheimer’s symptoms involve vocabulary reduction and repetitive speech, which are difficult to detect clinically at initial stages. Traditional diagnosis depends on visible symptoms and cognitive testing, often resulting in late detection. Recent research emphasizes the importance of identifying biomarkers in subtle physiological signals, prompting innovations like wearable sensors that can monitor these markers continuously and noninvasively.

“Analyzing body sounds with in-ear devices allows us to detect early physiological changes associated with neurodegenerative diseases before symptoms become apparent.”

— Professor at ÉTS

“Separating overlapping signals like speech, heartbeat, and breathing is challenging, but machine learning techniques are helping us decode these subtle cues.”

— Research team member

What Remains Unclear

It is not yet clear how accurately these signals can predict disease onset or progression across diverse populations. The effectiveness of the technology in real-world clinical settings remains to be validated through larger-scale studies.

What’s Next

The ongoing studies aim to validate the approach with larger patient cohorts and refine algorithms for better accuracy. Future steps include integrating these sensors into everyday devices and establishing protocols for clinical use.

Key Questions

How early can this technology detect neurodegenerative diseases?

It is currently under investigation; preliminary results suggest detection of subtle physiological changes months or years before clinical symptoms appear, but more research is needed to confirm exact timelines.

Can this in-ear device replace traditional diagnosis methods?

Not yet. It is intended as a supplementary tool to aid early detection, which can then be confirmed through standard clinical assessments.

Is this technology safe and comfortable for daily use?

Yes, the devices are designed to be lightweight and unobtrusive, similar to hearing aids or earbuds, suitable for continuous wear.

Will this technology be available to the public soon?

It is still in research phases; commercial availability will depend on further validation and regulatory approval, which could take several years.