You are currently viewing AI Smartphone App Transforms Ear Infection Diagnosis

AI Smartphone App Transforms Ear Infection Diagnosis

5/5 - (1 vote)

Physician-scientists at UPMC and the University of Pittsburgh have created a new cellphone app that uses artificial intelligence (AI) to precisely diagnose ear infections in children. The app, discussed in a recent article in JAMA Pediatrics, aims to reduce unnecessary antibiotic use in young children.

New AI smartphone tool accurately diagnoses ear infections

Ear infections, known as acute otitis media (AOM), are one of the most common reasons for prescribing antibiotics in children. However, accurately identifying AOM can be challenging without specialized training. The newly developed AI tool diagnoses ear infections by analyzing a short video of the ear drum, captured using an otoscope connected to a cellphone camera. This method offers a simple and effective solution, potentially outperforming trained clinicians.

Senior author Dr. Alejandro Hoberman emphasizes the common misdiagnosis of AOM, leading to inadequate care or unnecessary antibiotic treatment, which can impact the effectiveness of available antibiotics. The AI tool assists in obtaining accurate diagnoses and guiding appropriate treatment.

Approximately 70% of children experience an ear infection before their first birthday. Despite its prevalence, accurately diagnosing AOM requires expertise to identify subtle visual cues from a brief view of the ear drum, especially in infants. AOM is often confused with otitis media with effusion, a condition involving fluid behind the ear that typically does not require antibiotic treatment.

To enhance diagnostic accuracy, the research team built a training library of 1,151 videos of the ear drum from children who visited UPMC pediatric offices. Two experts reviewed the videos, diagnosing AOM or non-AOM. Using AI, two models were trained on features like shape, position, color, and translucency of the tympanic membrane.

The AI models demonstrated high accuracy, with sensitivity and specificity values exceeding 93%, surpassing previous clinician studies. Dr. Hoberman envisions the tool as a potential gamechanger in primary healthcare, supporting clinicians in stringent AOM diagnoses and treatment decisions.

Beyond accuracy, the tool allows captured videos to be stored in a patient’s medical record, facilitating sharing with other providers. It also serves as a valuable teaching tool for medical students, residents, and parents, aiding in understanding diagnoses and treatment decisions.

Dr. Hoberman anticipates widespread implementation of the technology in healthcare provider offices, improving AOM diagnoses and treatment decisions. The study received support from the Department of Pediatrics at the University of Pittsburgh School of Medicine.

Source: eurekalert