Like dirty diapers and sleepless nights, ear infections are just one of those things that come with having a young child. Sometimes the problem is obvious, complete with ear-tugging and screaming. Other times it’s not so obvious — leading to unnecessary doctor’s visits.
Researchers at the University of Washington have a clever fix for this problem. They’ve developed a smartphone app that can tell when fluid has built up behind the eardrum and predict the likelihood of an infection.
“Designing an accurate screening tool on something as ubiquitous as a smartphone can be game-changing for parents as well as health care providers in resource-limited regions,” said Shyam Gollakota, a co-author on the study and an associate professor of computer science and engineering at UW, in an announcement.
In the end, the app was successful in identifying the presence of fluid around 85 percent of the time, roughly in line with the accuracy of a specialist. The results of the study were published today in Science Translational Medicine.
The researchers are building a startup called Edus Health that will sell the app to the public.
In order to use the app, you have to get a bit crafty and create a paper funnel that attaches to the speaker and microphone. The phone then emits soft “chirps” that bounce off the eardrum and are recorded by the microphone. Ears full of fluid reflect the sound differently than healthy ears do, so the researchers trained a machine learning algorithm to detect the difference. The app assesses the likelihood of infection and recommends whether or not to see a clinician.
In addition to Gollakota, co-authors on the study included UW surgical resident Dr. Sharat Raju, Dr. Randall Bly, and doctoral students Justin Chan and Rajalakshmi Nandakumar.
The researchers tested their tech on 53 kids at Seattle Children’s Hospital, around half of whom were scheduled for a surgery in which a tube is placed into the ear to drain fluid. Those patients were tested before and after the surgery with the diagnostic app.
It’s the latest example of UW researchers using smartphones to diagnose health conditions. The university spun out Sound Life Sciences in February, a company that aims to commercialize an app that uses sonar to detect when an opioid user’s breathing has slowed. Gollakota also worked on that project and now serves as Sound’s CEO and CTO.
Separately, UW spinout Senosis health turned smartphones into monitoring devices that collect health metrics to diagnose pulmonary function, hemoglobin counts and other health information. Senosis was acquired by Google in 2017, and the deal was later revealed to be part of a digital health push by Google’s Nest smart home group.