Remote Acquisition of Vital Signs Using Infrared and Depth-Enabled Cameras
Ryan Chu (1), Matthew Downing (2), Apoorva SRivastava (2), ALEX Mihailidis (1,3,4), Ofer levi (1,2)
Institute of Biomaterials and Biomedical Engineering, University of Toronto
Department of Electrical and Computer Engineering, University of Toronto
Department of Occupational Science and Occupational Therapy, University of Toronto
Toronto Rehabilitation Institute, University Health Network, Toronto, ON
Vital signs are often used to provide an overall assessment of a subject’s physical health, but conventional methods of vital sign monitoring in hospitals are uncomfortable, non-portable, and expensive. We have developed a preliminary optics-based system of acquiring heart rate and respiratory rate, two of the most important vital signs, in real-time using cameras equipped with infrared sensors and 3-D vision. These cameras are portable, inexpensive, and enable non-contact physiological measurements. Infrared light projectors and sensors enable recovery of the blood volume pulse (BVP) from a recording of a subject, which is used to obtain heart rate. Multiple facial ROIs (forehead, upper lip, cheeks, etc.) are considered and results are compared to a clinical standard ECG/PPG. Respiratory rate is measured using an intensity map acquired from 3-D depth measurements. Use of 3-D depth information also improves performance with regard to subject motion detection and compensation. Applications of this new approach include routine clinical use for in-patient monitoring, as well as non-intrusive at-home monitoring for elderly and/or high-risk individuals. Future work will include testing of facial detection/recognition algorithms for tracking subjects within the camera’s field of view, as well as establishing hardware connections between multiple cameras to improve field of view and spatial resolution.