‘Selfie’ App Detects Pancreatic Cancer By Looking Into Your Eyes
Smartphones have revolutionized more than just the way we communicate: we now use them to read books, pay our bills, and even monitor our health. Mobile technology has become the future of medicine with the development of apps that can detect various health problems before they become visible. Researchers at the University of Washington have created an app that could detect the early signs of pancreatic cancer by simply looking at our eyes.
The app, “BiliScreen,” uses a smartphone camera, computer vision algorithms and machine learning tools to detect increased bilirubin levels — reddish yellow pigment formed by the breakdown of red blood cells — in a person’s sclera, which is the white part of the eye. Jaundice, the yellow discoloration of the skin, causes a rise in bilirubin levels before it’s visible to the naked eye. However, researchers found BiliScreen is able to detect color changes in the eye as people take a selfie.
In the initial clinical study, the app, used in conjunction with a 3D printed box that controls light exposure for the eye, accurately identified cases of concern 89.7 percent of the time in 70 people compared to the current blood test doctors use. This is significant because changes in sclera are normally detected once bilirubin levels are well past cause for concern.
Researchers used BiliScreen by using a smartphone’s built-in camera and flash, which collects photos of a person’s eye when they take a selfie. A computer vision system is able to automatically and effectively separate the white parts of the eye, which can be helpful for making a diagnosis. The app will then calculate the color information from the sclera; this is based on the wavelengths of light being reflected and absorbed, and correlate it with bilirubin levels using machine learning algorithms.
Aside from a 3D printed box, the team also tested the app with paper glasses printed with colored squares. This was used to help calibrate color. Yet, the researchers found using the app with the 3D printed box led to the best results.
Dr. Jim Taylor, a professor in the UW Medicine Department of Pediatrics, whose father died of pancreatic cancer at age 70, believes this technology shows promise for helping people who have pancreatic cancer catch it in time to have surgery and improve their odds of survival. Currently, pancreatic cancer has a five-year survival rate of 9 percent.
“Pancreatic cancer is a terrible disease with no effective screening right now,” said Taylor, in a statement.