A new mobile app developed by physician-scientists at UPMC and the University of Pittsburgh that uses artificial intelligence (AI) to accurately diagnose ear infections or acute otitis media (AOM) could help reduce of unnecessary antibiotic use in young children, according to new research published today in JAMA Pediatrics.
AOM is one of the most common childhood infections for which antibiotics are prescribed, but can be difficult to distinguish from other ear conditions without intensive training. The new AI tool, which makes a diagnosis by evaluating a short video of the eardrum captured by an otoscope attached to a cellphone camera, offers a simple and effective solution that could be more accurate than trained clinicians.
Acute otitis media is often misdiagnosed. Underdiagnosis leads to inadequate care and overdiagnosis leads to unnecessary antibiotic therapy, which can compromise the effectiveness of available antibiotics. Our tool helps in the correct diagnosis and guides the correct treatment.”
Alejandro Hoberman, MD, senior author, professor of pediatrics and director of the Division of General Academic Pediatrics at Pitt’s School of Medicine and chair of UPMC Children’s Community Pediatrics
According to Hoberman, about 70% of children have an ear infection before their first birthday. Although this condition is common, the accurate diagnosis of AOM requires a trained eye to detect subtle visual findings resulting from a brief projection of the ear drum in a dirty baby. AOM is often confused with otitis media with effusion or fluid behind the ear, a condition that generally does not involve bacteria and does not benefit from antimicrobial therapy.
To develop a practical tool to improve the accuracy of diagnosing AOM, Hoberman and his team began by creating and annotating a training library of 1,151 tympanic membrane videos from 635 children who visited UPMC pediatric outpatient clinics between 2018 and 2023. Two trained experts with extensive experience in AOM research reviewed the videos and diagnosed AOM or non-AOM.
“The eardrum, or tympanic membrane, is a thin, flat piece of tissue that runs along the ear canal,” Hoberman said. “In AOM, the eardrum swells like a bagel, leaving a central area of depression that looks like a bagel hole. In contrast, in children with otitis media with effusion, there is no bulging of the tympanic membrane.”
The researchers used 921 videos from the training library to train two different artificial intelligence models to detect AOM by examining features of the tympanic membrane, including shape, location, color and translucency. They then used the remaining 230 videos to test the models’ performance.
Both models were highly accurate, producing sensitivity and specificity values greater than 93%, meaning they had low false-negative and false-positive rates. According to Hoberman, previous studies of clinicians have reported diagnostic accuracy of AOM ranging from 30% to 84%, depending on the type of health care provider, level of education, and age of the children examined.
“These findings suggest that our tool is more accurate than many clinicians’,” Hoberman said. “It could be a game-changer in primary care settings to support clinicians in the rigorous diagnosis of AOM and in guiding treatment decisions.”
“Another advantage of our tool is that the videos we capture can be stored in a patient’s medical record and shared with other providers,” said Hoberman. “We can also show parents and trainees—medical students and residents—what we see and explain why we do or don’t diagnose an ear infection. It’s important as a teaching tool and to reassure parents that their child is receiving appropriate treatment.”
Hoberman hopes their technology could soon be widely implemented in healthcare providers’ offices to aid in the accurate diagnosis of AOM and support treatment decisions.
Other authors on the study included Nader Shaikh, MD, Shannon Conway, Timothy Shope, MD, Mary Ann Haralam, CRNP, Catherine Campese, CRNP, and Matthew Lee, all from UPMC and the University of Pittsburgh. Jelena Kovačević, Ph.D., of New York University. Filipe Condessa, Ph.D., Bosch Center for Artificial Intelligence. and Tomas Larsson, M.Sc, and Zafer Cavdar, both of Dcipher Analytics.
This research was supported by the Department of Pediatrics, University of Pittsburgh School of Medicine.
Source:
Journal Reference:
Shaikh, N., et al. (2024). Development and validation of an automated classifier for the diagnosis of acute otitis media in children. JAMA Pediatrics. doi.org/10.1001/jamapediatrics.2024.0011.
