Real-time remote photoplethysmography using a webcam and graphical user interface

Goal: This work is part of my 2019 summer research internship at Rice University in the Department of Electrical and Computer Engineering in partnership with the Rice Office of STEM Engagement.

Abstract: The pulse is an important vital sign that gives the medical state of a person. Access to affordable and reliable ways to measure vital signs is a struggle for many underserved communities. One solution is to create methods of measuring a pulse, or photoplethysmogram (ppg), using a camera from an existing device, like a smartphone or laptop. In this way, patients can connect to care-givers using telemedicine and their currently available devices. Telemedicine means patients can have remote access to care-givers in a more flexible way than traditional treatment pathways. Many devices exist for obtaining a ppg such as a pulse oximeter, but these devices require contact with the patient and are not easily integrated into telemedicine systems. The remote ppg technique allows for real-time analysis of the pulse using only a webcam. The remote ppg algorithm implemented here uses a user-selectable region of interest (roi) for which an intensity time-series is obtained. This project demonstrates a graphical user interface running on a computer for obtaining, analyzing, and visualizing the ppg waveform in real-time. Live signals of the remote ppg and a traditional pulse sensor are compared for ground truth analysis. The OpenCV and PyQtGraph packages in Python were used to create a relatively simple and robust way to obtain, analyze, and visualize the ppg in an engaging and accessible format which could be run locally or remotely. Future work would have the algorithm track the roi as the person moves in the frame.

Date: 1 June 2019