Scientists have created an app that uses Smartphone selfies to detect pancreatic cancer while it is in the early stages.
One of the earliest signs of pancreatic cancer is jaundice − a yellowing of the skin and eye sclera caused by an accumulation of bilirubin in the blood. By time yellowing of the sclera is visible to the human eye, the disease is often too advanced for treatment to be effective.
Researchers from the University of Washington conducted a study to see whether computer vision and machine learning tools could be used to detect signs of jaundice early on, before it is visible to the human eye and when bilirubin is only slightly elevated. Being able to do so could enable a new screening program for at-risk individuals.
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,"
Lead author Alex Mariakaki
The app, which is called BiliScreen uses a Smartphone’s camera and flash to take pictures of the eye when a person takes a selfie. A computer vision system then isolates the sclera and the app uses machine learning algorithms to correlate color information from the sclera with bilirubin levels.
When the team tested the app in an initial study of 70 people, BiliScreen correctly identified at-risk people almost 90% of the time, compared with the blood test healthcare professionals currently use to measure bilirubin levels. The blood test is only used when there is cause for concern because patients are already symptomatic. It also needs to be carried out by a healthcare professional and is not convenient for routine screening.
BiliScreen on the other hand, is a non-invasive, easy to use tool that people can use to check whether they need further testing. The app could also be used by individuals who already have pancreatic cancer and need to monitor their bilirubin levels.
BiliScreen is described in a paper that will be presented on September 13th at Ubicomp 2017, the Association for Computing Machinery's International Joint Conference on Pervasive and Ubiquitous Computing.
Next, Mariakakis and team plan to test the app on a wider range of people at risk of jaundice and make further improvements that will enhance usability.