An experimental AI-powered mesh sleeve can detect subtle signs of frailty among the elderly, researchers say.

The soft mesh sleeve, worn around the lower thigh, monitors a senior’s walking patterns, researchers recently reported in the journal Nature Communications.

An AI program contained within the sleeve then interprets that data to estimate frailty — the loss of strength, speed and energy that puts a senior at increased risk for falls, injuries and hospitalization.

“Right now, we often wait for a fall or hospitalization before we assess a patient for frailty,” said senior researcher Philipp Gutruf, associate head of biomedical engineering at the University of Arizona in Tucson.

“We wanted to shift the paradigm from reactive to preventative,” Gutruf said in a news release. "This device allows clinicians to intervene early, potentially preventing costly and dangerous outcomes."

About 15% of Americans 65 and older are affected by frailty, researchers said in background notes.

The 2-inch-wide mesh sleeve is lined with tiny sensors and “designed to be invisible,” Gutruf said.

The sleeve simultaneously records and analyzes the wearer’s leg acceleration, symmetry and step variability, researchers said.

Final results of the analysis are then transferred via Bluetooth to a smart device.

The continual AI analysis keeps the device from devouring power or stacking up tons of data, researchers said.

"Continuous, high-fidelity monitoring creates massive datasets that would normally drain a battery in hours and require a heavy internet connection to upload," lead researcher Kevin Kasper, a biomedical engineering doctoral candidate at the University of Arizona, said in a news release.

Tests among two small groups of seniors — one with 16 folks, another with 14 — found that the mesh sleeve worked about as well as current gold-standard ways of assessing frailty like grip strength, gait examination and sit-to-stand tests.

Researchers also had some participants wear the sleeve for 10 days and found that it could capture and analyze data over such an extended period.

The AI-enabled technology is "an ideal solution for remote patient monitoring in rural or under-resourced communities," Kaspar said. "We are effectively putting a lab on the patient, no matter where they live."

However, the device requires more testing among larger groups of patients before it will be ready to use, researchers said.