Variability in Heart Rate Estimates from Different Pulse Oximeters

J.Y.A. Foo, S.J. Wilson, G.R. Williams, C. Dakin, M.-A. Harris, and D.M. Cooper (Australia)


Data and signal acquisition, photoplethysmography, motion artefacts, pulse oximetry, and heart rate variability


Since its introduction, pulse oximetry has become a conventional clinical measure. Besides being arterial blood oxygen saturation (SpO2) measure, pulse oximeters can be used for other cardiovascular measurements, like heart rate (HR) estimations, derived from its photo plethysmographic (PPG) signals. The temporal coherence of the PPG signals and thereby HR estimates are heavily dependent on its minimal phase variability. A Masimo SET Rad-9TM, Novametrix Oxypleth and a custom designed PPG system were investigated for their relative phase variation. R-R intervals from electro-cardiogram (ECG) were recorded concurrently as reference. PPG signals obtained from the 3 systems were evaluated by comparing their respective beat-to-beat (B-B) intervals with the corresponding R-R estimates during a static test. For their relative B-B comparison to the ECG, Novametrix system differed 0.680.52% (p<0.05) and custom device 0.890.71% (p<0.05) while Masimo device 4.173.51% (p>0.05). Recent advanced signal processing approach is attractive with respect to the accuracy and noise immunity of the SpO2 measure. However, one consequence may be the variability it induces to the phase coherence of its PPG signal. Clinical interpretation of other measures such as HR may be compromised by these techniques.

Important Links:

Go Back