Estimating instantaneous respiratory rate from the photoplethysmogram

Area of Science:

• Biomedical Engineering
• Signal Processing
• Physiological Monitoring

Background:

• Photoplethysmogram (PPG) signals reflect blood volume changes and are influenced by respiration.
• Variations in venous return during breathing cause baseline shifts in PPG.
• Accurate estimation of instantaneous respiratory rate (IRR) is crucial for patient monitoring.

Purpose of the Study:

• To develop and validate an algorithm for estimating IRR from PPG signals.
• To leverage the synchrosqueezing transform (SST) for enhanced time-frequency analysis of PPG.
• To expand the diagnostic capabilities of pulse oximetry.

Main Methods:

• Applied the synchrosqueezing transform (SST) to PPG signals.
• Identified IRR as the dominant ridge within the respiratory frequency band (0.1-1 Hz) in the SST.
• Validated the algorithm against the Capnobase benchmark dataset, comparing with capnography and expert labels.

Main Results:

• The algorithm achieved a median RMS error of 0.39 breaths/min.
• Bland-Altman analysis indicated good agreement with reference respiratory rates (bias -0.32).
• The method demonstrated accurate IRR estimation from PPG data.

Conclusions:

• The proposed SST-based algorithm accurately estimates IRR from PPG.
• This technique enhances the functionality of pulse oximeters for non-invasive monitoring.
• IRR extraction from PPG offers valuable additional diagnostic information.