A New Way to Measure Vitals
Regular monitoring of vital signs is essential to anyone’s health and well being. To this extent, the search for easy and non-invasive methods to easily track vital signs is just as important. To date, vital signs are measured largely through contact sensors such, as electrocardiogram probes, pulse-oximeters, chest straps, and blood pressure cuffs—all of which are not ideal for day-to-day monitoring, and they are especially troublesome for the elderly or for newborn babies.
However, advanced algorithms are making it possible to estimate vital signs based on subtle changes in skin color, and we can do this using computer vision based on ordinary CMOS cameras or even mobile phone cameras.
Ultimately, these algorithms are are able to track almost imperceptible changes in a person’s skin color that can be enhanced to determine a diagnosis. It works because these color changes are caused by cardio-synchronous variations in the blood just under the surface of the skin.
How This Works
In theory, the technology has already existed for decades, but an effective implementation technique has yet to be found, thanks to a number of different challenges. For example, darker tones make it hard to determine small color changes, low lighting condition cause disruptions in diagnosis, and natural movement typically causes noise that corrupts the signal of the subject to the machine.
That said, now, these challenges have been addressed by new method that rely on a “divide and conquer” method to tackle the problems.
Researchers use a method that tracks different parts of the face independently that is pegged against a metric tracked separately. In their press release, the team outlines their new development, and how it functions after measurements are taken: “We then compute a ‘goodness metric’ for each tracked part and subsequently combine the signals in an adaptive way that is based on our defined metric. The steps involved in our algorithm are illustrated [below].”
The method offers an exciting breakthrough for clinical and wellness applications. And researchers behind the study believe that fusing the designs of illumination, camera optics, and signal processing algorithms will continue to improve the performance of implementation.
They assert, “By solving challenges in non-contact vital sign monitoring under motion, we believe that the accuracy of other tissue imaging techniques…can also be improved, as these modalities are also affected by motion artifacts.”