Using thousands of baby photos, some off-the-shelf software and the help of his colleagues, an assistant professor of chemistry and biochemistry at Baylor University is on a personal mission to develop a better way to detect early signs of pediatric eye cancer.
NPR today revisited the story of Bryan Shaw, who became a first-time father in 2008 to a son, Noah. Shaw’s wife is a photography enthusiast and naturally was always snapping photos of her new baby. But she noticed in some of her photos that there was a white glow coming from her infant son’s pupils.
With the Rise of AI, What IP Disputes in Healthcare Are Likely to Emerge?
Munck Wilson Mandala Partner Greg Howison shared his perspective on some of the legal ramifications around AI, IP, connected devices and the data they generate, in response to emailed questions.
During a checkup when the baby was 4 months old, she mentioned it to the doctor, who looked in Noah’s eyes and referred him to a specialist. The specialist made the diagnosis: a rare eye cancer called retinoblastoma, in both eyes.
That white glow that was appearing in the photos was leukocoria, which can be a sign of several eye problems in children. About half of the time, it’s caused by retinoblastoma.
Noah went through radiation therapy, chemotherapy and eventually eye surgery to remove his right eye. Throughout that heartbreaking process, Bryan Shaw’s inner scientist was seeking some answers. He began looking through thousands of old photos to see if Noah’s cancer could have potentially been caught earlier.
Using Photoshop and standard color space conversion algorithms, he detected the first signs of leukocoria in digital pictures of his son at 12 days old. He also noticed that the eye with the larger tumors, the right eye, was more white in pictures.
In collaboration with some computer science colleagues, oncologists and the ophthalmologist who treated Noah at the Massachusetts Eye and Ear Infirmary, Shaw developed a quantitative scale of leukocoria that he believes could potentially help parents and doctors identify retinoblastoma sooner.
A prototype version of the leukocoria detection software the team developed is available for free online, but it “comes with no guarantees.” To turn it into a commercializable test, they would need to do more thorough clinical testing to show that it works and can distinguish between retinoblastoma cases and false positives.
Retinoblastoma is usually not fatal in the U.S., but diagnosis often comes later in developing countries. Shaw has said such a test could be especially helpful in countries where access to digital photography will increase at a faster rate than access to pediatric eye exams.
[Image credit: Baylor Communications]