Remote sensing of physiological signs using a machine vision system. (4th July 2017)

• Record Type: Journal Article
• Authors: Al-Naji, Ali; Gibson, Kim; Chahl, Javaan
• Is Part Of: Journal of medical engineering & technology. Volume 41:Number 5(2017)
• Page Start: 396
• Page End: 405
• Publication Date: 2017-07-04
• Subjects: Long-distance monitoring -- colour magnification technique -- motion magnification technique -- imaging photoplethysmography -- face detection -- graphic user interface; Biomedical engineering -- Periodicals; Medical technology -- Periodicals; 610.28
• DOI: 10.1080/03091902.2017.1313326
• Languages: English
• ISSNs: 0309-1902

Abstract

Abstract: The aim of this work is to remotely measure heart rate (HR) and respiratory rate (RR) using a video camera from long range (> 50 m). The proposed system is based on imperceptible signals produced from blood circulation, including skin colour variations and head motion. As these signals are not visible to the naked eye and to preserve the signal strength in the video, we used an improved video magnification technique to enhance these invisible signals and detect the physiological activity within the subject. The software of the proposed system was built in a graphic user interface (GUI) environment to easily select a magnification system to use (colour or motion magnification) and measure the physiological signs independently. The measurements were performed on a set of 10 healthy subjects equipped with a finger pulse oximeter and respiratory belt transducer that were used as reference methods. The experimental results were statistically analysed by using the Bland–Altman method, Pearson's correlation coefficient, Spearman correlation coefficient, mean absolute error, and root mean squared error. The proposed system achieved high correlation even in the presence of movement artefacts, different skin tones, lighting conditions and distance from the camera. With acceptable performance and low computational complexity, the proposed system is a suitable candidate for homecare applications, security applications and mobile health devices.