Taylor–Fourier Analysis of Blood Pressure Oscillometric Waveforms

Abstract

Blood pressure oscillometric waveforms behave as amplitude modulated nonlinear signals with frequency fluctuations. Their oscillating nature can be better analyzed by the digital Taylor-Fourier transform (DTFT), recently proposed for phasor estimation in oscillating power systems. Based on a relaxed signal model that includes Taylor components greater than zero, the DTFT is able to estimate not only the oscillation itself, as does the digital Fourier transform (DFT), but also its derivatives included in the signal model. In this paper, an oscillometric waveform is analyzed with the DTFT, and its zeroth and first oscillating harmonics are illustrated. The results show that the breathing activity can be separated from the cardiac one through the critical points of the first component, determined by the zero crossings of the amplitude derivatives estimated from the third Taylor order model. On the other hand, phase derivative estimates provide the fluctuations of the cardiac frequency and its derivative, new parameters that could improve the precision of the systolic and diastolic blood pressure assignment. The DTFT envelope estimates uniformly converge from K=3, substantially improving the harmonic separation of the DFT.