Researchers at Brown University have developed an algorithm to determine if someone has anemia by looking at their picture on a smartphone.
Anemia affects about a third of the people in the world. It is a disease where patients have insufficient red blood cells; hence they cannot get enough oxygen to their body tissues. If not treated, anemia can lead to heart problems, complications for pregnant women, and fatigue. It can even lead to death in the elderly, children, and chronically ill people. About 12% of children less than five years old and 15% of women age 15-49 have anemia.
Custom-made smartphones can usually be designed with attachments that allow someone to take a blood-based test for anemia. However, the method is complicated as it requires many additional extensions, extra costs and the need for testing equipment outside the phone.
Usually, anemia is tested for at hospitals and clinics. There is also a kit available that allows you to take the test yourself. However, this new test gives patients in isolated areas or countries with limited access to healthcare an easy way to take the test themselves.
According to Dr. Selim Suner, a study author, people can take pictures of the vascular surface of their lower eyelids to predict anemia. She adds that this will meet the need for affordable, non-invasive and accessible ways to diagnose anemia. When someone has anemia, their palpebral (a part of their lower eyelid) becomes pale.
The researchers collected smartphone images of the palpebral of 142 patients. Using these photographs, they created an algorithm that compared the color of the palpebral with the surrounding skin of the eye. They tested the new tool on 202 different patients and managed to diagnose more than 70% of the cases accurately. The algorithm also helped them predict whether patients needed a blood transfusion.
The patient’s skin tone did not affect the results. However, poor quality images did not yield accurate results.
As a result of the study, people who suspect they have anemia could use their smartphones for an affordable, non-invasive test instead of going to the clinic for a blood test.