University of Pittsburgh scientists have developed an artificial intelligence-based phone application they say accurately diagnoses ear infections.
They even contend that the app could help reduce dependence on antibiotics, which are often over-used in dealing with childhood ear problems due to misdiagnosis.
The app could be a "gamechanger" as it appeared to be "more accurate than many clinicians," according to Alejandro Hoberman, professor of paediatrics at the University of Pittburgh's School of Medicine.
Hoberman and colleagues said they used 921 videos to "teach" AI models to "detect AOM by looking at features of the tympanic membrane, including shape, position, colour and translucency."
The app, which is used to to assess "a short video of the ear drum captured by an otoscope connected to a cellphone camera," could prove more accurate than doctors, who do not always have the "intensive training" needed to discern infections such as acute otitis media (AOM), which is treated with antibiotics, from other common ailments, such as fluid behind the ear, which are usually not bacterial.
As any bleary-eyed parent knows, ear problems, including infections, are common in babies, with around 7 in 10 affected before their first birthday, leading to sleepless nights.
The app could in turn provide a small boost in the fight against antimicrobial resistance, which has been growing due to the overuse or misuse of antibiotics, not only by people against conditions where they have no function, but on a bigger scale as part of farm animal feeding programmes in countries such as China.
In 2021, England’s former chief medical officer warned that antimicrobial resistance could prove "the death knell for modern medicine." Annual worldwide deaths linked to this have been estimated at between half a million and 6.5 million people worldwide. Last year, the European Centre for Disease Control and Prevention warned that e.Coli was becoming resistant to antibiotics.
