3D printed smartphone add-on uses selfies to spot cancer

A screening tool for pancreatic cancer has been developed by researchers at the University of Washington.

By utilising smartphone technology and 3D printing the researchers have developed a solution that could effectively screen for pancreatic cancer and other diseases, simply by snapping a selfie.  

Pancreatic cancer has one of the worst survival rates, with just 1% of adults expected to survive past 10 years from their initial diagnosis, according to Cancer Research UK. The disease is difficult to diagnose in its early stages as symptoms rarely show until the cancer has spread.

Jaundice, a yellow discolouration of the eyes and skin, is caused by an increased level of bilirubin in the blood. Whilst jaundice is one of the earlier signs of pancreatic cancer, it is still only visible to the naked eye when in severe stages.

The solution, BiliScreen, assesses the levels of jaundice in a person’s eyes when they take a selfie. The image is then tested using computer vision and machine learning to assess if the person has a mild form of jaundice.

The researchers propose BiliScreen as a solution to detecting pancreatic cancer at an earlier stage. In a 70-person clinical study, the researchers used the app in conjunction with two accessories that were designed to limit the effects of external lighting, resulting in a clearer picture. As described in the University of Washington study, the accessories include a ‘3D-printed box that controls the eyes’ exposure to light and paper glasses with coloured squares for calibration’.

When a selfie is taken, the app isolates the white parts of the eye and measures the colouration of the sclera. Machine learning algorithms then assess the eye’s bilirubin levels by looking at the wavelengths that are being reflected and absorbed.

Results of the study showed that the 3D printed box, used in conjunction with the smartphone app, was the best option for identifying accurate levels of bilirubin. Compared to standard blood tests the study showed that the app was able to accurately identify cases of concern 89.7% of the time.

Lead author of the study, Alex Mariakakis, said: “The problem with pancreatic cancer is that by the time you’re symptomatic, it’s frequently too late. 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.”