AN ADVANCED NONLINEAR SIGNAL MODEL TO ANALYZE PULSATION-DERIVED PHOTOPLETHYSMOGRAM SIGNALS

This paper proposes an advanced nonlinear signal model to handle the intrinsic nonlinearity of the pulsation of photoplethysmogram (PPG) signals and so permit their assessment. This model consists of three-different sinusoidal source signals and it successfully reproduces the primary harmonic, the second harmonic, the respiratory and the other spectra stemming from the interaction of the three source signals. We demonstrate the potential of the model in various experiments on PPG signals. It is discovered that the model has a great performance in that it characterizes PPG signals captured under various measurement conditions if nonlinearity is small. It is also demonstrated that the model successfully characterizes the post-alcohol-intake state and the post-physical-exercise state using the behavior of the nonlinear factors, which should be very useful for alcohol-intake detection. The remaining issues of the model are also addressed.