AUTHOR=Mejía-Mejía Elisa , Kyriacou Panicos A. TITLE=Spectral analysis for pulse rate variability assessment from simulated photoplethysmographic signals JOURNAL=Frontiers in Physiology VOLUME=Volume 13 - 2022 YEAR=2022 URL=https://www.frontiersin.org/journals/physiology/articles/10.3389/fphys.2022.966130 DOI=10.3389/fphys.2022.966130 ISSN=1664-042X ABSTRACT=Pulse rate variability (PRV) refers to the changes in pulse rate through time and is extracted from pulsatile signals such as the photoplethysmogram (PPG). Due to the widespread use of PPG sensors in many healthcare and wellbeing settings, PRV has been used as a surrogate of heart rate variability (HRV), which is measured from the electrocardiogram (ECG) and reflects the changes of heart rate through time. Nonetheless, HRV and PRV have been shown to have differences, and it has been hypothesised that these differences may arise from physiological processes or from technical aspects that may affect the reliable extraction of PRV from PPG signals. Moreover, there are no guidelines for the extraction of PRV information from pulsatile signals, which hinders the comparison among PRV studies and the understanding of physiological changes that may affect PRV. In this study, the extraction of frequency-domain information from PRV was studied, in order to establish the best performing combination of parameters and algorithms to obtain the spectral representation of PRV. Using simulated PPG signals with known PRV content, it was found that using fast Fourier transform and the multiple signal classification (PMUSIC) algorithms gave the best results, combined with cubic spline interpolation and a frequency resolution of 0.0078 Hz for the former; and a linear interpolation with a frequency resolution as low as 1.22 x 10^(-4), as well as applying a fifth order model, for the latter. Hence, and considering the lower complexity of FFT over PMUSIC, FFT should be considered as the appropriate technique to extract frequency-domain information from PRV signals.