• Posted March 5, 2024

New AI Tool Could Diagnose Ear Infections

A new smartphone app can help doctors more accurately diagnose ear infections in babies and small children, potentially reducing unnecessary antibiotic use in kids, researchers report.

Ear infections -- known by doctors as acute otitis media (AOM) -- can be difficult to diagnose, as they require a trained eye to detect subtle clues from a brief view of the eardrum of a wriggly baby, researchers said.

About 70% of children have an ear infection before their first birthday, researchers said.

However, it's easy for doctors to mistake simple fluid behind the ear -- a condition called otitis media with infusion -- for an actual ear infection, researchers said. That condition does not involve bacterial infection and doesn't benefit from antibiotics.

"Acute otitis media is often incorrectly diagnosed,"said senior researcher Dr. Alejandro Hoberman, director of the division of general academic pediatrics at the University of Pittsburgh School of Medicine. "Under-diagnosis results in inadequate care and over-diagnosis results in unnecessary antibiotic treatment, which can compromise the effectiveness of currently available antibiotics."

To improve the accurate diagnosis of ear infections, the research team compiled a training library of 1,151 videos from 635 children treated at the University of Pittsburgh Medical Center.

All the videos involved examination of the tympanic membrane in the ear, Hoberman said.

"The eardrum, or tympanic membrane, is a thin, flat piece of tissue that stretches across the ear canal,"Hoberman explained in a university news release. 

The researchers then used 921 videos from the library to teach two different AI models to detect ear infections by examining the features of the tympanic membrane.

"In AOM, the eardrum bulges like a bagel, leaving a central area of depression that resembles a bagel hole,"Hoberman said. "In contrast, in children with otitis media with effusion, no bulging of the tympanic membrane is present."

The team then tested the AI's ability to detect ear infections, using the remaining 230 videos from the library.

Both AI models were fairly precise, telling ear infections from fluid behind the ear with better than 93% accuracy.

By comparison, previous studies have found that the accuracy of human doctors ranges between 30% and 84%, researchers said.

The new study was published March 4 in the journal JAMA Pediatrics.

"These findings suggest that our tool is more accurate than many clinicians,"Hoberman said. "It could be a gamechanger in primary health care settings to support clinicians in stringently diagnosing AOM and guiding treatment decisions."

The examination videos from each patient also could be stored with their medical record, helping doctors manage future ear problems, Hoberman added.

"We can also show parents and trainees -- medical students and residents -- what we see and explain why we are or are not making a diagnosis of ear infection,"Hoberman said. "It is important as a teaching tool and for reassuring parents that their child is receiving appropriate treatment."  

More information

Johns Hopkins Medicine has more about ear infection.

SOURCE: University of Pittsburgh, news release, March 4, 2024