A smartphone app that uses artificial intelligence (AI) to help accurately diagnose acute otitis media could reduce unnecessary antibiotic use in young children, US researchers have reported.
The team developed two AI models to analyse a few seconds of video taken from an otoscope connected to a phone camera to detect acute otitis media by assessing features of the eardrum, including its shape, position, colour, and translucency, they reported in JAMA Pediatrics.1
These models were developed using 900 videos of eardrums from children aged under 3 who had visited their paediatrician from 2018 to 2023, and the team then tested them on 230 different videos. Both models were highly accurate, as the final decision support tool showed a sensitivity of 93.8% and specificity …
