Music In, Diagnosis Out

Right now, earbuds let you listen to your favorite tunes. Soon, they may also be monitoring the health of your ears.

EarHealth, developed by a UB-led research team, pairs standard earbuds with a smartphone that’s equipped with deep learning-based algorithms, a form of artificial intelligence. The system works by sending a series of chirps through the earbuds of a user and recording how the chirps reverberate throughout the ear canals. These initial recordings help create profiles of the individuals’ unique inner-ear geometry and physical characteristics under various conditions.
 
Subsequent chirps—for example, a user might set the system for once-daily testing—monitor each ear for three conditions that alter the ear canal’s geometry: earwax blockage, a ruptured ear drum and otitis media, a common ear infection.

Each condition has a unique audio signature that the deep learning-based algorithms can detect with fairly accurate results.

In a recent study, conducted with a diverse group of human subjects, researchers reported that the current iteration of EarHealth achieved an accuracy rate of 82.6% among 92 users, including 27 healthy subjects, 22 with a ruptured eardrum, 25 with otitis media and 18 with earwax blockage. The study was partially supported by the U.S. National Science Foundation.

“With people worldwide living longer, and the prevalence of headphones, it is more important than ever to monitor one’s ear health,” said corresponding author Zhanpeng Jin, associate professor in the Department of Computer Science and Engineering, School of Engineering and Applied Sciences.

“With EarHealth, we have developed what we believe is the first-ever earbud-based system that monitors ear health conditions in an effective, affordable and user-friendly way,” he added. “Because it can be used in people’s daily lives and has the potential to detect these conditions very early, it could greatly improve health outcomes for many people.”