At the Johns Hopkins Data Science and AI Institute‘s fall symposium, AI in Daily Life, held September 16 on the Homewood campus and open to the public, engineers, computer and data scientists, medical researchers, and clinicians presented what they’re doing to contend with pressing health-system challenges.
Presenting at the symposium was Mathias Unberath, DSAI’s director of research, interactive, and embodied AI and a John C. Malone Associate Professor of Computer Science with joint appointments in ophthalmology and otolaryngology at the School of Medicine.
“Seeking medical care [in the U.S.] usually comes with a ton of frustration,” Unberath told the crowd. “You don’t get the care where you need it—in your home—and you have to [make arrangements] to secure child care and travel to your doctor’s office, assuming your provider even has an open appointment.”
Unberath described the common scenario of waking up with a sore throat and not knowing whether it was strep, the contagious bacterial infection treated with antibiotics, or something like a seasonal cold that needed to run its course and would likely resolve on its own. “You wouldn’t want to spread strep at work or school,” he said. “And you’d want to feel better, so you’d need to know.”
To help, Unberath and his colleagues worked with Therese Canares, Bus ’21 (MBA), a pediatric emergency physician and entrepreneur, to develop a mobile app that serves as a strep-throat screening tool. Users simply upload a photo of their throat. The app then compares the photo to AI images and determines the likelihood of strep in a matter of seconds, offering guidance on self-care strategies and whether to see a doctor.
Dubbed CurieDx, the app is named after the two-time Nobel Prize-winning Polish-French scientist Marie Curie—a decision made by Canares, who put her clinical work on hold to bring the app to market and add more screening capabilities, say, for rashes and eye infections. Currently, doctors and nurses can use CurieDx in the clinic and via telehealth to help diagnose strep throat, and people everywhere will soon be able to use it at home or on the go, Unberath said.
According to Unberath, AI-based tools like this hold the potential to transform the medical system—in particular, by easing the burden on the health system at a time when chronic illnesses are rising, our population is aging, and physician and nurse shortages are leaving patients without the care they need, he said. Specialists like neurologists, psychiatrists, cardiologists, ophthalmologists, and others are especially in short supply, with patients sometimes waiting more than a year for an appointment, Unberath told the audience.
AI can help, he and others contended. For example, another AI-based device Unberath is developing equips general eye-care providers with the training and tools they need to detect glaucoma, a progressive condition and a leading cause of blindness. Typically, ophthalmologists who specialize in glaucoma make the diagnosis, but securing an appointment often involves a long wait time. With this new AI tool, general eye doctors will be able to spot glaucoma, gauge the severity, and start or manage a treatment regimen, while referring patients with fast-progressing or severe conditions to a specialist.