A multiple signal classification approach for photoplethysmography signals in healthy and athletic subjects

In this paper, various spectral analysis methods are applied to investigate the spectral behaviour of Photoplethysmography (PPG) signals for both healthy normal athletic and non–athletic subjects. The paper investigates the contribution of the PPG spectral components in revealing time trends and how they minimise artefacts that conceal these spectral contents. Both parametric and non–parametric spectral methods are utilised to study the PPG signal. However, eigenstructure methods such as multiple signal classification are found to be more suitable to localise spectral contents of the signal and robustly estimate the heart rate. The heart rate detection correlation is found to be 99.62% and the PPG detection accuracy is 99.61%. These findings show that PPG can be used as an effective tool in investigating cardiovascular disorders as well rather than in investigating only blood oxygenation and heart rate measurements. Furthermore, it can advance the research for existing heart rate variability methods.