A new app from Baylor University takes advantage of a smartphone's camera -- and of the smartphone owner's tendency to take a lot of pictures -- to potentially detect rare eye cancers in babies.
Baylor chemist Bryan Shaw and Baylor computer scientist Greg Hamerly have launched the White Eye Detection app, in which parents can upload pictures of their babies and digitally scan them for signs of rare eye diseases including retinoblastoma, pediatric cataract, and Coats' disease.
In conditions like these, tumors in the back of the eye that can't be seen normally can show up in digital photographs as a white pupil in one or both of the child's eyes. As NPR reports, Shaw's own son had retinoblastoma and he was able to detect it after noticing white eye -- properly called leukocoria -- in a baby photo.
"If I would have had some software telling me, 'Hey, go get this checked out,' that would have sped up my son's diagnosis and the tumors would have been just a little bit smaller when we got to them," Shaw told NPR. "There might have been fewer."
The app can automatically look through all the baby pictures on an iPhone or iPad's camera roll and flag any potential leukocorias, which is important because they tend to show up only inconsistently in photographs. When it finds pictures, it suggests that parents go to their pediatrician. It also requests that they upload the photos to Baylor's database, so they can be used to continue refining the algorithm. Ideally, the app will learn to be better and better at eliminating false positives that might cause needless worries for patients.
An additional feature is a screening mode, which allows the user to shine the phone's light into someone's eye and use the camera to search for the telltale white reflection.
The algorithm was developed out of work done on photographs of Shaw's son and eight other children for a paper published last year in PLOS ONE. The paper established that a large number of photographs taken in aggregate could even give information about the size of the tumor and the progression of the disease.