A group of researchers in Canada and China are working on a revolutionary new method to monitor blood pressure: a selfie.
The groups, based at the University of Toronto and the Affiliated Hospital of Hangzhou Normal University, published a proof of concept that enables smartphone customers to track blood pressure by taking a short video.
“We’re utilizing a tech known as transdermal optical imaging,” Kang Lee, a developmental psychologist at the Toronto University and the lead author of the paper informed Quartz. The idea is to make use of light emitted by a smartphone camera to detect proteins because it bounces off the skin. Utilizing machine learning algorithms, the group can then translate these measurements into blood pressure readings, readings they say are accurate about 95 % of the time.
However, there’s one major caveat: the test group of 1,328 was large of East Asian or European descent, meaning the proof-of-concept might be wildly inaccurate in folks of color.
In concept, the test should work the same means. In testing fairer skin, the test is wanting at the measurements of light bouncing off protein like hemoglobin. It’s unclear if it’ll run into issues with greater instances of another protein, melanin, in those with darker skin. It should be capable of informing the difference; however, without testing on individuals with varying skin tones, it’s impossible to say if the test might maintain its high levels of accuracy.
Lee and his team acknowledged that the researchers needed to test the method on other skin types.