In a prospective proof-of-concept study, researchers developed a supervised machine learning model to differentiate otitis media with effusion from normal tympanic membranes using smartphone-captured otoscopic images, achieving 81% diagnostic accuracy with 87% sensitivity and 74% specificity on internal testing. These findings suggest that smartphone-based imaging combined with machine learning may support more accurate frontline diagnosis of middle ear effusion, particularly in pediatric and primary care settings, although external validation is needed.

Source: American Journal of Otolaryngology