image: Dr. Randall Bly, an assistant professor of otolaryngology-head and neck surgery at the UW School of Medicine who practices at Seattle Children's Hospital, uses the app and funnel to check his daughter's ear. This material relates to a paper that appeared in the May 15, 2019, issue of Science Translational Medicine, published by AAAS. The paper, by J. Chan at University of Washington in Seattle, WA; and colleagues was titled, "Detecting middle ear fluid using smartphones." view more
Credit: Dennis Wise/University of Washington
Scientists have created a user-friendly smartphone-based platform that can quickly detect the presence of fluid in the middle ear - a likely indicator of ear infections - in children. The new system, which was validated in 98 patient ears in a pediatric surgical center, could provide a low-cost and effective tool for parents to detect ear infections such as acute otitis media (AOM), a leading cause of visits to pediatric health providers. Some cases of AOM can lead to severe complications such as meningitis, and other pediatric ear conditions such as otitis media with effusion are associated with delays in speech and poorer school performance. Detecting middle ear fluid remains a critical need in pediatric medicine, but existing tests such as tympanometry cannot be performed at home or require expensive equipment. To overcome these hurdles, Justin Chan and colleagues created a system that uses a commercial smartphone's existing speakers and microphones to send audible "chirps" of sound into the ear canal. A machine learning algorithm then analyzes the reflected acoustic signals and predicts middle ear status (a deeper pitch in the sound indicates that the middle ear is filled with pus or fluid). The system requires no additional equipment beyond a small paper funnel to focus the sound, which can be constructed from printer paper, scissors and tape in under three minutes. The researchers tested the system using an iPhone 5S and a Galaxy S6 on 98 ears from patients between 18 months and 17 years of age, and found that the platform detected middle ear fluid as well as or better than established techniques such as acoustic reflectometry. Importantly, parents of patients operated the system as easily as physicians in a separate experiment involving 25 patient ears, and the software was not substantially affected by background noise such as the sound of a crying child.
###
Journal
Science Translational Medicine