A new smartphone app which uses artificial intelligence can accurately diagnose acute otitis media (AOM) by connecting to an otoscope, a study has found.
The free iPhone app from the University of Pittsburgh assesses a short video of the patient’s eardrum captured by connecting an otoscope to the smartphone camera.
Senior author Dr Alejandro Hoberman, a paediatrics professor, said the automated classifier interpreted videos of the tympanic membrane to enhance accuracy in OAM diagnosis and assist treatment.
He said study findings suggested it was “more accurate than many clinicians” and another benefit was videos captured could be stored in a patient’s medical record and shared with other providers.
Physician scientists from the University of Pittsburgh and the University of Pittsburgh Medical Centre developed the Pitt CMU iTM app with engineers and AI specialists from America and Sweden.
They conducted a study of its use in children to determine if it could be used in a primary care setting to enhance diagnostic accuracy.
The study, published in JAMA Pediatrics, concluded it was highly accurate, producing sensitivity and specificity values over 93%, meaning low rates of false negatives and false positives. Using 1,151 videos of the tympanic membrane from 635 children at outpatient clinics it found sensitivity of 93.8% and specificity of 93.5%.
“These findings suggest that given its high accuracy, the decision-support tool could reasonably be used in primary care or acute care settings to aid with decisions regarding treatment of acute otitis media,” the researchers wrote.
Hoberman is professor of paediatrics and director of the Division of General Academic Pediatrics at Pitt’s School of Medicine and president of UPMC Children’s Community Pediatrics.
“Underdiagnosis results in inadequate care and overdiagnosis results in unnecessary antibiotic treatment, which can compromise the effectiveness of currently available antibiotics,” he said. “Our tool helps get the correct diagnosis and guide the right treatment.”
Hoberman said about 70% of children had an ear infection before their first birthday but accurate diagnosis required a trained eye to detect subtle visual findings gained from a brief view of the eardrum on a wriggly 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 treatment.”
Library of over 1,000 videos
To develop the tool, the researchers built and annotated a training library of 1,151 videos of the tympanic membrane from 635 children who visited outpatient clinics. Two experts with extensive experience in AOM research reviewed the videos and made a diagnosis of AOM or not AOM.
“In AOM, the eardrum 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,” Hoberman said.
The researchers used 921 videos from the training library to teach two different AI models to detect AOM by looking at features of the tympanic membrane including shape, position, colour and translucency. They then used the remaining 230 videos to test how the models performed.
Both models were highly accurate, producing sensitivity and specificity values over 93%.
Hoberman said previous studies of clinicians had reported diagnostic accuracy of AOM ranging from 30% to 84%, depending on the health care provider, level of training and age of the children being examined.
“These findings suggest that our tool is more accurate than many clinicians,” said Hoberman. “It could be a gamechanger in primary health care settings to support clinicians in stringently diagnosing AOM and guiding treatment decisions.”
He said health care providers using the app could show parents and medical trainees what was seen and explain why they are or are not making a diagnosis of ear infection.
“It is important as a teaching tool and for reassuring parents that their child is receiving appropriate treatment,” he said.
An editorial said the app was “a rare example of carefully developed and highly accurate AI/ML (machine learning) models designed for paediatrics and ear, nose, and throat specialties”. Of 692 AI/ML-enabled medical devices approved by the FDA, only two were for use in the ENT field.
