Effect of Number of Modes of EMD in Respiratory Rate Estimation from PPG Signal

Abstract

Hertzman invented photoplethysmography, a non-invasive electro optic technology that delivers information on the blood volume flowing at a specific area on the human body near the skin. Multiple attempts on PPG Derived Respiration have been made; these approaches are based on various signal processing strategies such as wavelets, filtering, and other statistical methods. The principal component analysis is a technique for identifying patterns in the dataset and expressing them in a way that highlights their similarities and contrasts. As the patterns in data might be difficult to discover in high-dimensional data without the benefit of graphical presentation, PCA is a useful technique for evaluating such data. Empirical Mode Decomposition is suitable for extracting key components that are specific to the underlying biological or physiological processes. This paper examines the EMD method and associated algorithms, as well as some examples of applications. The suggested EMD technique successfully retrieved respiratory information from PPG signals when tested on PPG signals from the well-known Capnobase database from the Physio bank archive. Moreover, the method's superiority was demonstrated by the evaluated similarity metrics in both the time and frequency domains for original and predicted respiratory rates.