A team of researchers from the Computer Science and Artificial Intelligence Laboratory at the Massachusetts Institute of Technology has developed an algorithm that can measure one’s heart rate through videos by analyzing the small and imperceptible movements of the head.
These head movements, which are caused by blood flow occurring during heart contractions, are difficult to see with the naked eye but can be detected using video amplification.
Tests using the system showed that the algorithm yielded pulse measurements that were nearly as exact as that of measurements produced by standard electrocardiograms or EKGs.
This system of measuring pulse can potentially be useful in monitoring newborns or the elderly whose sensitive skin could suffer from the constant attachment and removal of EKG patches. The system can also potentially help in identifying patients who are at risk of cardiac-related issues by monitoring the volume of blood pumped by the heart.
Graduate student Guha Balakrishnan and his advisors John Guttag and Fredo Durand discuss their research in a paper coming out this summer at the Institute of Electrical and Electronics Engineers’ Computer Vision and Pattern Recognition conference.
Watch the video demo of the technology below.