Decomposing photoplethysmogram waveforms into systolic and diastolic waves, with application to hyperbaric environments

Abstract

Objective:A new algorithm to decompose the photoplethysmogram (PPG) pulse in two waves related with the systolic and diastolic parts, was applied to identify alterations in the morphology of the PPG pulse due to the pressure. Approach:Each pulse was decomposed into two waves: the first one a Gaussian related with the systolic peak, and the second one was modelled as a lognormal curve associated with the diastolic part. From these two waves, 13 parameters related with the width (W1, W2 and W2/W1), the time instant (T1, T2, T21, TBB), the amplitude (A1, A2, A2/A1) and the areas (D1, D2, D2/D1) were estimated. These parameters were computed from subjects inside a hyperbaric chamber, involving five stages at different pressure: 1 atm, 3 atm, 5 atm, 3 atm and 1 atm. Main results:There was a significant increase in the values of A1, A2, W1, T1, and D1, and a decrease in the ratios when the pressure increased, that suggest a vasoconstriction when the pressure increased. There is also an increase in the values of TBB, T2, T12, and W2 along the protocol, that implies a dependency of some parameters with the pulse-to-pulse interval. Significance:This methodology allows extracting a large set of parameters related with the PPG morphology that are affected by the change of pressure inside hyperbaric environments.