Some exceptional students at the University of Washington have just come up with an application that can tell if you have acute pancreatic cancer or not by just looking at the whites of your eyes.
BiliScreen is a mobile application that takes the patient’s selfie using a smartphone’s camera. It then uses it to identify the disease before it reaches the dangerous level. The application runs complex Computer Vision and Machine Learning algorithms on the image to come up with a thorough diagnosis. The application gauges increased levels of bilirubin in the patient’s eye whites to detect earlier stages of pancreatic cancer, like the setup of Jaundice.
The application will be presented in a conference due to take place 13th of September. During the testing phases, the application was able to identify the disease correctly 89% of the times. The students are using this application with a VR type headset for better lighting conditions leading to better accuracy.
Pancreatic Cancer is malignant disease known for its exceptionally low survival rates. Through the course of past 5 years, only 9 out of a hundred patients have known to survive it. The reason for this low survival rate is often being credited to the late diagnosis of the fatal disease. Now, these students at the University of Washington have come up with an application that can diagnose a pancreatic cancer in its initial stages even before the symptoms in the eye-whites become visible to the humans.
Source — University of Washington