Your child’s ear infection? There’s an app for that.

Ear infections are a common scourge of childhood. Five out of six children have at least one before they turn three.

Now, a new AI-enabled smartphone app can diagnose pediatric ear infections. And it’s potentially more accurate than clinicians’ current diagnostic protocols.

The University of Pittsburgh researchers behind the tool hope their technology can help prevent inappropriate treatment for this common condition. Let’s dig into how the app would do that.

Image: Pexels
Image: Pexels

Over- or under-treatment

The most common type of ear infection is acute otitis media (AOM). It consists of a middle ear infection where swelling traps fluid behind the eardrum, causing pain and sometimes fever.

However, AOM is often incorrectly diagnosed. 

This can lead to undertreatment, when the condition, for instance, has progressed past the acute stage. Or, the condition may be overtreated. In this second misdiagnosis scenario, a child suspected of having AOM may in fact have otitis media with effusion—fluid behind the ear—which is generally not a bacterial condition and thus does not require antibiotics.

“Underdiagnosis results in inadequate care and overdiagnosis results in unnecessary antibiotic treatment, which can compromise the effectiveness of currently available antibiotics,” said senior author Alejandro Hoberman. “Our tool helps get the correct diagnosis and guide the right treatment.”

Besides ensuring children get the right care, this public health concern is one of the biggest reasons it’s crucial to not misdiagnose such a common bacterial infection. If antibiotic overuse leads to resistance, our current drugs may no longer work, leading a common, treatable infection to turn into something more dangerous for more patients.

Currently, experts estimate over a quarter of antibiotics prescribed in outpatient settings are inappropriate. And over-eager telehealth prescription is thought to be exacerbating the issue.

So, how can doctors and parents be confident their child is being correctly diagnosed and treated? For AOM? There’s an app for that.

Differentiating between ear conditions

Previous studies of clinician diagnostic accuracy in this area estimate that providers correctly identify ear infections 30–84% of the time.

The ear infection diagnostic app, in contrast, has specificity and sensitivity at or above 93%. 

How does it work?

A clinician operating the tool takes a short video with a cellphone camera connected to an otoscope and the app analyzes it, looking for the telltale signs of AOM.

“The ear drum, or tympanic membrane, is a thin, flat piece of tissue that stretches across the ear canal,” Hoberman said. “In AOM, the ear drum bulges like a bagel, leaving a central area of depression that resembles a bagel hole. In contrast, in children with otitis media with effusion, no bulging of the tympanic membrane is present.”

Why this difference in accuracy?

As it turns out, visually diagnosing an ear infection is no walk in the park. Clinicians often only get a brief look into a young (often wriggly) child’s ear to search for the subtle cues indicating AOM. 

Hoberman’s team used 921 videos from a 1,151 video training library to teach two models to perform this visual identification. They then used the remaining videos to test performance.

Plus, beyond assuring accurate treatment, the video taken of a patient’s ear can be stored in their electronic health record (EHR). From there, it can be used to assure parents of the clinician’s diagnosis —and teach medical students and residents about proper ear infection diagnosis.

A simple tool with a wide array of benefits—from ensuring kids get the right care to preventing antimicrobial resistance. That’s the kind of elegant health innovation we love to see.

Read more

See all

MedTech Pulse is a newsletter publication on innovation at the intersection of technology and medicine. Stay ahead with unique perspectives on industry news, the latest startup deals, infographics, and inspiring conversations.

Powered by