View the video
A little jiggle of the head may be all that doctors need to measure pulse.
A new computer algorithm can detect the nods, bobs and wiggles people make as blood rushes in and out of their skulls, and then translate the motions into heart rate.
Results from the motion-sensing technique are nearly identical to those from an electrocardiogram, a test that uses electrodes stuck to the skin to pick up signals from the heart, report MIT computer scientist Guha Balakrishnan and colleagues. They will present the work June 27 at the Computer Vision and Pattern Recognition conference in Portland, Ore.
Using a video, the researchers’ algorithm tracks up to 1,000 points on a person’s face, zeroes in on the regular twitches that accompany blood pumping to the head and then calculates heart rate from the timing of the twitches. Though the motions are barely perceptible by eye, Balakrishnan amplifies the movements to show people’s heads bouncing and shimmying like bobblehead dolls.
Doctors could use the algorithm as a hands-off way to measure the pulses of patients with delicate skin, such as babies and elderly people. The technique might also spot heart blockages, the researchers suggest. If blood flows unevenly through the heart, the head may shake in a tell-tale way.
Using a computer algorithm to track barely perceptible head movements, MIT researchers can measure a person’s heart rate from a video of their face.
Credit: Guha Balakrishnan, Fredo Durand, John Guttag/ MIT
G. Balakrishnan et al. Detecting pulse from head motions in video. Presentation at the Computer Vision and Pattern Recognition conference in Portland, Ore., June 27, 2013.
Note: To comment, Science News subscribing members must now establish a separate login relationship with Disqus. Click the Disqus icon below, enter your e-mail and click “forgot password” to reset your password. You may also log into Disqus using Facebook, Twitter or Google.