PLAN 80: Article
The Hands-Free Heart Monitor

Using Low-Cost Video to Measure Vital Signs

A device developed by the Affective Computing group at the Media Lab has been named one of the ten best inventions of 2011 by Popular Science Magazine; it has also won a prize of $50K in a competition sponsored by the Center for Integration of Medicine and Innovative Technology, an organization created by doctors and engineers to develop new tools to meet clinical needs.

Developed by Ming-Zher Poh, (SM'07, EECS) a 2011 graduate of the Harvard-MIT Health Sciences and Technology program, working with Daniel McDuff, a graduate student in the Media Arts and Sciences program and their advisor Rosalind Picard, Head of the Affective Computing group, the project was published last year in the peer-reviewed journal Optics Express.

The invention, called Cardiocam, is a non-contact technology for measuring physiological signals such as heart rate and breathing rate using a simple low-cost camera such as the webcam in a laptop computer.

Such noninvasive monitoring could allow doctors to check the vital signs of burn victims or babies without attaching uncomfortable clips. It could also be used for telemedicine screening over the Internet using a patient's own webcam or cell-phone camera.

A system like this could also be built into a bathroom mirror so that patients who need ongoing monitoring, or people who simply want to keep track of their health, could get routine readings displayed in a corner of the mirror.

Such a mirror is actually now on display at the Media Lab: behind the two-way glass, a webcam-equipped monitor is wired to a laptop so that when you stand in front of the mirror the monitor projects your heart rate on top of your reflection.

The technology is based on light readings. Every time your heart beats, it sends a pulse of blood through your blood vessels; blood absorbs light, so when a surge of it travels through the vessels less of the light that hits your skin is reflected.

The team developed an algorithm that can pick out the small fluctuations in the heart rate's light pattern from all the other reflected light captured by a webcam, then wrote code to process the data in real time, allowing the laptop to generate an instant heart-rate reading.

The big development challenge was dealing with movements of the subject and variations in ambient lighting. But the team was able to adapt signal-processing techniques originally developed to extract a single voice from a roomful of conversations to extract the pulse signal from the 'noise' of these other variations.

The system is able to generate valid results even when the subject is moving around a bit in front of the camera and produces pulse rates that agree to within about three beats per minute with the rates obtained from traditional monitoring devices. It is even able to get accurate pulse signals from three people in the camera’s view at the same time.

Researchers have tracked this effect with a high-res camera in the past, but the use of a simple webcam makes nearly every computer and smartphone a potential heart-rate monitor. In time, the system may be able to measure other vital signs as well, including respiratory rate and blood-oxygen saturation.