An Optimal Reflection Photoplethysmographic Sensor System Based on Skin Optics

This paper aims at optimizing the placement of sensors in reflection photoplethysmographic sensor system by analyzing the roughness of surfaces of skin layers and estimation of optimum position of the optode pair for photoplethysmograph (PPG). The rough surfaces are assumed to have Gaussian or Rayleigh surface auto covariance (ACV) functions, and their theoretical bidirectional reflectance distribution functions (BRDFs) are calculated using the Rayleigh-Rice surface scatter theory. Measurement of BRDF is used to characterize the scattering properties of surface. A spectrophotometer arrangement consisting of light-emitting diodes and photodetectors has been assembled and tested for wavelengths 470, 650, and 935 nm to measure the BRDF of surfaces of skin layers. From the experimental analysis, the surface ACV function of the skin layers is obtained based on the measured BRDF. The experimentally obtained BRDFs are compared with the theoretical measurements of Gaussian and Rayleigh ACV functions. The results indicate that the skin surfaces follow Rayleigh ACV function with ACV length (lc) equal to 0.3 times the wavelength of incident radiation. The root-mean-square value of surface roughness (σs) for each surface of skin is calculated from the measured reflected light. From the experimental results obtained for BRDF, the direction of the scattered light is analyzed, and thereby, the position of optode pair for reflection PPG is found.