Non-contact Monitoring Of Ambulatory Blood Pressure Based On Webcam

Measurements of ambulatory blood pressure can provide a vital physiological assessment. However, attempts so far are contact and non-automated, susceptible to external factors and typically expensive. In this paper, a non-contact method of measuring ambulatory blood pressure is presented. This approach allows for efficient pulse rate acquisition directly from face image captured from webcam and is based on imaging PhotoPlethysmoGraphic. In essence, given the two pulse signals, the phase difference and the heart rate can be obtained. The desired signal was acquired by the functional relationship between phase difference, heart rate, and blood pressure. The proposed algorithm is thoroughly assessed on a dataset containing 249 videos. Using Bland-Altman for correlation analysis, we compared blood pressure extracted from video recorded by a basic webcam with the sphygmomanometer and achieved high accuracy and correlation. Furthermore, the proposed technique can measure more than one person at a time and may have a great value in monitoring a person at home.