Cardiopulmonary signal acquisition from different regions using video imaging analysis

Abstract

Activity of the cardiovascular and respiratory systems causes some imperceptible physiological and physical effects in different regions of the human body, which can be highly informative in clinical and biomedical settings, where physiological signs and lifestyle indicators are required to be measured remotely. The possibility of extracting cardiopulmonary signals from many different ROIs (face, forehead, head, palm, wrist, arm, neck, leg and chest), based on cost effective systems without any biological effects on humans, has been investigated in this paper using different image and signal processing techniques. The experimental results of all proposed systems (the colour-based method and the motion-based method) approved their effectiveness, high correlation coefficients (PCC = 0.98 for cardiac measurements and 0.93 for respiratory measurements) and low error rates (RMSE = 1.97 beats/min for cardiac measurements and 1.98 breaths/min for respiratory measurements) in comparison with those obtained from conventional reference methods. Therefore, the proposed systems in this paper may be of value for upcoming non-contact monitoring systems in the diagnostic and health care applications.