Estimation of Respiratory Rate from Photoplethysmogram Signal Using Maximal Overlap Discrete Wavelet Transform

Respiratory Rate (RR) abnormality is often an early indicator of critical illness. Reliable estimation of RR continuously from other physiological signals is an active area of research. Direct measurement of RR can be done by measuring oronasal airflow. However, this is an uncomfortable process for the patient. Hence, conventional approaches use the electrocardiogram (ECG) signal to analyze respiratory sinus arrhythmia. Photoplethysmogram (PPG) waveform analysis provides an alternate, non-invasive method of estimation, which is suitable for patients of all ages and health conditions. In this work, various techniques available in the literature regarding RR estimation from the PPG waveform have been studied. A methodology based on Maximal Overlap Discrete Wavelet Transform (MODWT) is proposed for Multi-Resolution Analysis (MRA) to provide the RR estimate. Comparison of the proposed technique with the existing algorithms demonstrates better accuracy for estimating RR.