Scattering-driven PPG signal model

Abstract

This article discusses the origin of photoplethysmographic (PPG) signals. Two plausible hypotheses that could explain the phenomenon of pulsatile optical signals are analyzed: the volumetric hypothesis in which changes in the arterial blood volume are responsible for the observed signal pulsation, and a model wherein changes in the measured signal are driven by the size fluctuations of red blood cell (RBC) aggregates. The theoretical approximation where the size of scattering particles representing RBC aggregates varies as a function of pulsatile changes in blood flow is elaborated. Within this models framework, the gamma coefficient used in pulse oximetry was calculated for the volumetric-related and aggregation-related models. Two pairs of wavelengths, (670 nm, 940 nm) and (590 nm, 940 nm), were selected to determine gamma. As a function of aggregate size, the gamma behavior was simulated for these pairs and the two hypotheses. To verify the model predictions experimentally, the PPG signals at the fingertip were measured using reflection geometry. Two combinations of light-emitting diodes with two pairs of wavelengths were utilized as light sources. To manipulate the length of aggregates in the blood, external pressure was applied to the fingertip, presumably reducing the blood flow velocity. The gamma values were determined. The derived results fully agree with the theoretical predictions of the aggregation-driven PPG signal model. In addition, using a pressure sensor, the oscillometric signal amplitude in the fingertip and the PPG signal amplitude were simultaneously measured. The comparison results of oscillometric and optical signals at elevated external pressure values are not consistent with the volumetric hypothesis. All of the foregoing experimental results strongly support the argument favoring the incorporation of the proposed aggregation mechanism into the generic PPG signal model.