Mobile Health App Can Detect Ear Infections through Microphone

A group of researchers from the University of Washington have recently created a mobile health app that detect ear infections through the phone’s speaker and microphone. To work, all that is needed other than the smartphone is a funnel made from folded paper and the AI-powered app, developed by Justin Chan and colleagues at the University of Seattle. Their findings were published May 15 in Science Translational Medicine.

In this technique, the smartphone sends soft audible signals into the ear through the paper funnel, which are reflected to the phone’s microphone. Based on the frequency interpreted, the phone’s app determines how likely there is for fluid to be present in the ear (indicative of ear infections) with 85 percent accuracy. This percentage is comparable to those of diagnostic procedures used by specialists to detect fluid buildup in the middle ear, with common methods including acoustics or puffing air into the ear.

“Designing an accurate screening tool on something as ubiquitous as a smartphone can be game changing for parents as well as health care providers in resource limited regions,” said Shyam Gollakota, co-author of the study and associate professor in the UW’s Paul G. Allen School of Computer Science & Engineering. “A key advantage of our technology is that it does not require any additional hardware other than a piece of paper and a software app running on the smartphone.”

The app’s primary function is to send sounds into the ear, and measure how the sound waves are altered after interacting with the eardrum. The eardrum, formally known as the tympanic membrane, is the barrier between the outer and middle ear. If the middle ear is filled with fluid due to infection, these sound waves will reflect off the eardrum in a different manner than if they encountered a healthy ear. The smartphone app detects and analyzes these discrepancies in frequency and predicts the likelihood of fluid buildup.

READ MORE: How Will 5G Wireless Impact Healthcare?

“It’s like tapping a wine glass,” said co-first author Justin Chan, a doctoral student in the Allen School. “Depending on how much liquid is in it, you get different sounds. Using machine learning on these sounds, we can detect the presence of liquid.”

The funnel used can be made from any piece of paper that either the physician or consumer can cut and fold to form the tool. This funnel rests on the patient’s outer ear and guides the soundwaves to and from the eardrum. The 150-millisecond audio clip resembles a bird chirping and was found to be effective in the researcher’s work.

To train their machine learning algorithm, the researchers analyzed 53 children between 18 months to 17 years of age at the Seattle Children’s Hospital. Roughly half of the children were scheduled for surgical ear tube replacement to treat chronic fluid buildup, and the others were scheduled for other surgeries not related to the ear. Prior to their operation, each patient’s ear was analyzed using the smartphone system.

“What is really unique about this study is that we used the gold standard for diagnosing ear infections,” said co-first author Dr. Sharat Raju, a surgical resident at the UW School of Medicine. “When we put in ear tubes, we make an incision into the eardrum and drain any fluid present. That’s the best way to tell if there is fluid behind the eardrum. So these surgeries created the ideal setting for this study.”

In the ear tube group, surgery revealed fluid presence in 24 middle ears, and absence in the other 24. Among the participants scheduled for other operations, two eardrums showed symptoms of ear infection, while the other 48 appeared healthy. The machine learning algorithm successfully predicted the likelihood of fluid presence 85 percent of the time. The team then tested the algorithm on 15 ears, analyzing children aged between 9 and 18 months. All five ears that contained fluid were detected, and 9 of the 10 healthy ears were correctly identified.

“Fluid behind the eardrum is so common in children that there’s a direct need for an accessible and accurate screening tool that can be used at home or in clinical settings,” explained Raju. “If parents could use a piece of hardware they already have to do a quick physical exam that can say ‘Your child most likely doesn’t have ear fluid’ or ‘Your child likely has ear fluid, you should make an appointment with your pediatrician,’ that would be huge.”

Source: University of Washington

Jack holds a biology degree from Penn State University, and has a keen interest in how new medical technologies are changing the future of healthcare. Reach out to Jack if you have a compelling story idea or with feedback about past articles.