A team of researchers at Seattle-based University of Washington are analyzing smartphone selfies for symptoms of pancreatic cancer.
The standard screening method for pancreatic cancer involves an invasive blood test, which physicians tend to administer if there is a reason for concern, rather than on a routine basis. The UW research team is working to create a noninvasive screening method, based on one of the earliest symptoms of pancreatic cancer: jaundice, a yellow discoloration of the skin and eyes.
The team developed an app, called BiliScreen, which asks users to take a selfie using their smartphone's built-in camera. The app then leverages computer vision algorithms and machine learning tools to detect signs of jaundice in the whites of a person's eyes. Based on the app's findings, it suggests whether a patient should consult with a physician for further testing.
In an initial clinical trial with 70 participants, the researchers compared findings as determined by the BiliScreen app with those from the standard blood test. When the app was used in conjunction with a box that controlled the participant's exposure to light, it correctly identified 89.7 percent of cases the blood test flagged as concerning.
"The problem with pancreatic cancer is that by the time you're symptomatic, it's frequently too late," said lead author Alex Mariakakis, a doctoral student at the Paul G. Allen School of Computer Science & Engineering at UW. "The hope is that if people can do this simple test once a month — in the privacy of their own homes — some might catch the disease early enough to undergo treatment that could save their lives."