Respiratory Rate Estimation from PPG
Respiratory rate can be extracted from PPG through three modulation mechanisms: amplitude modulation (respiratory variation in pulse amplitude), frequency modulation (respiratory sinus arrhythmia in IBI), and baseline modulation (intrathoracic pressure changes shifting DC baseline). Fusion of all three provides 1–2 breaths/min accuracy.
Respiration modulates PPG through three physiological pathways. Amplitude modulation: inspiration reduces venous return to the left heart, decreasing stroke volume and pulse pressure, causing ~5–15% respiratory-synchronous variation in PPG AC amplitude. Frequency modulation: respiratory sinus arrhythmia causes IBI to shorten during inspiration and lengthen during expiration, creating 0.15–0.4 Hz oscillations in the IBI time series. Baseline modulation: intrathoracic pressure changes during breathing shift the venous pressure and tissue blood volume, causing ~1% respiratory-synchronous variation in the PPG DC component.
Each modulation source is extracted separately: amplitude modulation via peak envelope detection, frequency modulation via IBI series spectral analysis, and baseline modulation via low-pass filtering of the raw PPG. The respiratory frequency is estimated from each source using peak detection or spectral analysis in the 0.1–0.5 Hz band (6–30 breaths/min). Smart fusion combines estimates using quality-weighted averaging, where the weight for each source is proportional to its spectral peak prominence.
Wearable respiratory rate from PPG achieves MAE of 1–2 breaths/min during rest and 2–4 breaths/min during light activity vs. reference respiratory belt. During vigorous exercise, motion artifacts confound all three modulation sources, degrading accuracy to 3–8 breaths/min. Apple Watch, Garmin, and Fitbit all report respiratory rate during sleep, where accuracy is highest (MAE < 1.5 breaths/min).
Frequently Asked Questions
Which PPG respiratory modulation source is most reliable?
Frequency modulation (RSA in IBI) is most robust during rest. Amplitude modulation provides the strongest signal during mechanical ventilation. Baseline modulation is most susceptible to motion. Fusion of all three consistently outperforms single-source estimation.
Can PPG detect abnormal respiratory patterns?
PPG-derived respiratory rate can detect tachypnea (>20 breaths/min), bradypnea (<12 breaths/min), and Cheyne-Stokes patterns in heart failure. Detection of apnea events requires SpO2 desaturation monitoring rather than respiratory rate estimation alone.
What is the minimum PPG window for respiratory rate estimation?
30 seconds captures approximately 5 respiratory cycles at normal breathing rate and provides adequate spectral resolution. 60-second windows are standard for clinical validation studies.