AI is transforming how doctors diagnose ear infections

Ear infections are painful and can potentially damage hearing. Sometimes they're misdiagnosed or even overdiagnosed.

Now researchers are turning to artificial intelligence to help doctors get a better read on this common condition.

Acute otitis media or ear infections hurt and can lead to hearing loss and speech delays.

Five out of six kids will develop one by their third birthday.

The reason? Eustachian tubes connecting the ear to the back of the nose are underdeveloped, requiring artificial drainage tubes.

"They basically are ventilating the middle-ear space. So you create a hole in the ear drum, you place a tube in that's keeping that hole open, and that allows it to ventilate the middle ear and remove any fluid that might be behind the eardrum," said Dr. Aaron Moberly, professor of otolaryngology at Vanderbilt University Medical Center.

Otoscopes have been in use for decades, but lack proper lighting and clear field of view.

Moberly is now working with artificial intelligence to record and analyze eardrum videos.

"The software then takes that, creates a composite image, but then also provides some feedback back to the clinician and says, 'This is a 90% chance that this isn't a normal-looking ear, or this is a 75% chance of there's fluid behind the eardrum," he said.

The AI algorithm will use deep learning to identify and classify middle ear disease with greater accuracy. This should prevent long-term hearing loss and improve speech and language development.

While there are apps to diagnose ear infections, Moberly cautions against buying an otoscope online and using phone apps to self diagnose.