How GE is measuring blood pressure with selfies

Regular blood pressure measuring could soon be completely cuffless, thanks to a group of scientists and researchers who are working on creating algorithms to detect blood pressure using a smartphone camera.

Scientists at GE Global Research, Michigan State University and the University of Rochester Medical Center are working on making algorithms that can analyze video from a smartphone’s camera to detect the blood pulsing below the skin to read blood pressure.

“This new cuff less approach involves capturing a short video of your face and hands lasting 5 to ten seconds, during which we observe and analyze what’s happening beneath the skin to estimate your heart rate and blood pressure,” Lalit K. Mestha, an engineer at GE Global Research, said in a press release.

Natural light is able to penetrate just under the illumination under the skin, but the human eye cannot. The researchers report that the algorithm only needs several seconds of close-up footage of a person’s hands of face to detect pulse information. It uses the color shifts in the skin that is caused by changes in blood volume to put out a blood pressure reading.

“The color variation is generally too subtle for the naked eye to detect, but not for the algorithm,” Mestha said.

Blood pressure is typically measured using blood pressure cuffs that can cost anywhere from $40 to $75 for home use. People who are pregnant, people with high blood pressure and other conditions have to regularly take their blood pressure to ensure it’s not too high for too long. The smartphone app being developed by the researchers could help reduce the costs of using home blood pressure cuffs while allowing for continuous monitoring.

“Imagine being able to continuously monitor the vitals of entire neonatal units in the hospital or home monitoring for the elderly and disabled,” Gayu Seenumani, a senior engineer at GE Global Research, said. “For patients with high blood pressure who much take blood thinners for a heart condition, being able to regularly and more easily take their blood pressure would be much more convenient and could be a lifesaver.”

The American College of Cardiology and the American Heart Association recently changed the hypertension guideline to 130/80 mm of mercury or greater. This means that people with this blood pressure measurement are more at risk of having a stroke or heart attack than the previous 140/90 reading.

GE is hoping the app will help make healthcare more wireless and cloud-based.

“The world is going wireless and wearable,” Erno Muuranto, an engineer at GE Health Innovation Village in Helsinki, Finland, said. “We could run hospitals like smart factories. Wireless sensors and data analytics will help correctly diagnose patients in the ambulance. It will allow us to administer correct treatment faster, which could lead to faster discharge. It will also allow us to monitor people remotely from home. All of this will help improve care and costs.”

Researchers at MIT created an app last year that used similar light-based algorithms from selfies on a smartphone to detect heart arrhythmia. The app calculates heart rate based on facial light reflection. As a user’s heart beats, the volume of blood in a user’s vessels also increases. Hemoglobin in blood absorb’s light and decreases the light that is reflected by the skin. MIT’s app uses that knowledge to track the changes in the skin to calculate a heartbeat.

The researchers are currently testing the app on volunteers from the University of Rochester Medical Center.

The research was funded by the National Institutes of Health.