University of Arizona researchers based in the Gutruf Lab have developed a comfortable, easy-to-use wearable device that incorporates artificial intelligence to detect subtle warning signs of frailty, signifying a potential leap forward in elderly care.
Discussing the innovation, lead researcher Philipp Gutruf says: “The current model of care is lagging behind. Right now, we often wait for a fall or hospitalization before we assess a patient for frailty. We wanted to shift the paradigm from reactive to preventative.”
The project study introduces a soft mesh sleeve worn around the lower thigh that monitors and analyses leg acceleration, symmetry and step variability.
The result is a device framework that integrates artificial intelligence with clinical grade biosignal acquisition at the edge, performing on-device inference with clinical grade fidelity over extended durations with no interaction required by the wearer.
Frailty, which indicates greater susceptibility to falls, disabilities and hospitalization, affects 15% of U.S. residents 65 and older, according to a 2015 study in the Journals of Gerontology (“Frailty in Older Adults: A Nationally Representative Profile in the United States”).
Form and function define design
The researchers spent the last seven years developing technology that monitors biomarkers. His lab published a study in May on an adhesive-free wearable that measures water vapor and skin gases to track signs of stress.
Adapting and expanding on that technology, the approximately two-inch-wide, 3D-printed sleeve lined with tiny sensors is “designed to be invisible,” adds Gutruf.
The sleeve simultaneously records and analyses motion of the wearer and produces an AI analysis. With the device sending just the results, not the actual hundreds of hours of recorded data, transmission is reduced by 99% and the need for high-speed internet is eliminated.
Results are transferred via Bluetooth to a smart device. And long-range wireless charging capabilities free the user from plugging in the device or swapping out a battery.
Continuous, high-fidelity monitoring conventionally creates massive datasets that would normally drain a battery in hours and require a heavy internet connection to upload. The researchers solved this with Edge AI.
The research appears in the journal Nature Communications, titled “Wearable AI for on-device frailty assessment.”
