Release date: 2017-09-04
Pancreatic cancer is one of the worst prognosis cancers, with a five-year survival rate of only 9% because it has no significant signs of symptoms and a non-invasive screening tool to detect tumors before the cancer spreads.
Now, researchers at the University of Washington have developed an app that makes it easier to scan for pancreatic cancer and other diseases by taking a selfie on a smartphone.
Image source: University of Washington
Called BiliScreen, the app uses a smartphone camera, computer vision algorithms, and machine learning tools to detect elevated levels of bilirubin in the sclera or in the white of the eye. The app was described in an article by the Computer Society's Universal Computer International Conference Ubicomp 2017.
Image source: University of Washington
One of the early symptoms of pancreatic cancer is jaundice, which causes the skin and eyes to turn yellow due to the accumulation of bilirubin in the blood. Early signs of jaundice were found to be a small increase in bilirubin, but before the naked eye can be observed, this ability is a completely new screening procedure for at-risk individuals.
Image source: University of Washington
A total of 70 subjects were included in the initial clinical study of BiliScreen APP. Using a 3D printed box to control the exposure of the glasses to light, APP correctly identified 89.7% of the cases compared to current blood tests.
"The problem with pancreatic cancer is that once symptoms appear, it's too late," said Alex Mariakakis, a Ph.D. student in computer science and engineering at Paul G. Allen College, a research leader. "We want people to do it once a month." Such simple tests can be carried out in their own homes, some diseases are discovered early, and timely treatment is expected to save their lives."
BiliScreen's early work came from the Pervasive Computing Lab at the University of Wisconsin. They had previously developed BiliCam, a smartphone app that scans the baby's skin for scanning newborn jaundice. A recent study published in the journal Pediatrics showed that BiliCam provides an accurate assessment of bilirubin levels in 530 infants.
Image source: University of Washington
Working with medical doctors at the University of Wisconsin, UbiComp Labs can use professional equipment such as cameras, microphones and other components on mobile phones and laptops to make good use of them to screen for diseases.
Nowadays, doctors routinely use blood tests to measure bilirubin levels. Unless there are reasons for special concern, frequent examinations are not performed in adults, and blood tests need to be performed by professional medical institutions.
Image source: University of Washington
BiliScreen is designed to be very easy to use, as a non-invasive tool that can help confirm if a user needs to consult a doctor for further testing. In addition to diagnosis, BiliScreen can potentially alleviate the burden of pancreatic cancer patients who need to monitor bilirubin frequently.
In adults, whites can more easily reflect the level of bilirubin than skin, which can help early detection of signs of pancreatic cancer, hepatitis or generally harmless Gilbert syndrome. Unlike skin color, changes in the sclera can serve as a common standard across races.
By Shwetak Patel
"Eye is a very interesting portal for the body - tears can tell you how much glucose you have, and the sclera can tell you how much bilirubin is in your blood," said the Washington Research Foundation Entrepreneurship Award-authorized professor. Shwetak Patel, senior author of computer science and engineering and electrical engineering, said, "Our question is: Can we capture these changes and hopefully lead to the early detection of self-portraits (diseases)?"
Image source: University of Washington
BiliScreen uses the camera functions of smartphones and quickly collects photos of human eyes as they take selfies. The team developed a computer vision system that automatically and effectively distinguishes the white part of the eye, which is a valuable tool for medical diagnosis. The APP can then calculate the information of the scleral color based on the wavelength of the reflected and absorbed light and correlate it to the bilirubin level by a machine learning algorithm.
Image source: University of Washington
In order to calculate different lighting conditions, the team tested two different wearing modes of BiliScreen, one is the 3D printed box form, which can control the lighting conditions, and the other is the colored square glasses form, which can help the APP to correct the color. The use of the box mode app is reminiscent of Google Cardboard.
The research team will next conduct a wide range of APP tests on jaundice and potential risk groups to continuously improve usability, including the removal of unnecessary wearing devices such as boxes and glasses.
By Jim Taylor
"This relatively small range of primary research shows that the technology is still very promising," said Jim Taylor, a co-author of the University of Washington School of Medicine, whose father died of pancreatic cancer at age 70.
"Pancreatic cancer is a very terrible disease," Taylor said. "Our goal is to have more people who have unfortunately got pancreatic cancer. They can be lucky to find it in time and give them a better chance of survival."
Every time the scientists work hard, the god of death goes further.
Reference material
[1] New app could use smartphone selfies to screenfor pancreatic cancer
[2] A selfie and a smartphone app could catchpancreatic cancer early
Source: Health New Vision (Micro Signal HealthHorizon)
Brand computer chips, dual-frequency integrated circuits, cross-magnetic induction detection, bringing higher accuracy, reliability and stability. 2. Automatic control system, using special magnetic materials and unique technology, the detection accuracy of the probe is particularly high, and the accurate position of the broken needle is displayed in 8 detection areas. 3. Dual probe design, double detection function, higher safety performance, compared with other types of needle detectors, at the same detection height, the sensitivity is increased by 10%. 4. Automatic detection, automatic induction, automatic stop when no detection is needed, 0-10 sensitivity adjustment, when there is a broken needle, the conveyor belt automatically stops and returns backwards. 5. With a counting function, it can count the number of qualified or unqualified products. 6. It has two alarm modes: sound effect and light alarm. 7. Luxurious appearance, novel style, time-saving and power-saving, simple operation
Rapid PCR detection,Rapid Detection Rapid Detection PCR Kit Viral Antigen Detection
Jiangsu iiLO Biotechnology Co., Ltd. , https://www.sjiilogene.com