Non-contact method of blood pressure estimation using only facial video

摘要

In this paper, we investigate the possibility of using image-based pulse transit time (iPTT) to estimate blood pressure using an RGB (red, green, blue)-camera video taken from the face only. It is generally known that there is a correlation between iPTT and blood pressure. iPTT refers to the difference between the times when a pulse wave reaches two different parts of the body. Therefore, it is possible to estimate blood pressure without contact, with iPTTs obtained from the RGB camera. However, this method requires simultaneous capture of images of the face and palm. Recording of such videos is limited by the subject's posture. Hence, it is very difficult to achieve blood pressure estimation in a real-world environment. To solve these problems, this study examined whether it is possible to measure pulse wave propagation based on a single part of the body. The results showed that there was a time difference between the pulse waves at the forehead and the chin, thus it was possible to measure iPTT using only the face. Pulse waves were obtained from the chin and forehead, and the correlations between iPTT and blood pressure from the acquired pulse waves were examined. The results showed a negative correlation between these iPTT measures and blood pressure. The results suggest that simple, non-contact blood pressure estimation will be possible in the future.