Non Contact Measurement of Vital Signs

Overview

The purpose of this study is to test the accuracy of a web cam-based biomedical device developed at UVA (not FDA-approved) that is designed to measure heart rate, respiratory rate, and oxygen saturation without requiring any patient contact. One potential application of such a device would be in the field of infant monitoring allowing parents (and physicians) to monitor the vital signs of infants continuously. The investigators therefore propose to record the heart rate, respiratory rate, and oxygen saturation of 100 infants (defined as children aged 12 months or less) who are receiving continuous oxygen, heart rate, and respiratory rate monitoring with a traditional vital signs monitor. The relationship between "non-contact" and "gold standard" (GE monitoring equipment) heart rate, respiratory rate, and oxygen saturation will be analyzed using regression and limits of agreement analysis.

While multiple investigators have attempted to develop non-contact pulse oximeters, none of these devices have achieved accuracy sufficient for clinical use, no such devices have been approved by the Food and Drug Administration, and there are currently no such devices on the market in the United States. While these devices are typically able to measure the heart and respiratory rates with some accuracy,v the accurate calculation of oxygen saturation from the arterial pulse (SpO2) using a "non-contact" reflectance oximetry probe is complicated by the interference of ambient light, patient temperature changes, as well as the inherent limitations of the sensing devices currently utilized. Poh et al have been somewhat successful at calculating the heart rate from a video recording using independent component analysis. However, Poh's method does not calculate instantaneous rates and requires a facial recognition component to track the facial orientation in the image, is not capable of measuring respiratory rate, and relies primarily on analysis of reflected green light (which cannot be used for the calculation of oxygen saturation). Our work involves modifying a commercial off the shelf (COTS) 3-channel (red, green, blue) CCD (charge coupled device) or CMOS (complementary metal oxide semiconductor) camera in the form of a web cam to detect near infrared and infrared spectrum radiation and applying an algorithm based fast Fourier transformation (FFT) of individual red pixel intensity to detect motion and color changes. Because our algorithm analyzes the first derivative of red pixel intensity, a face-tracking component is unnecessary, and we are able to calculate the heart rate and the respiratory rate in real time.