Active Noise Cancellation filtering of motion-corrupted PPG data enabled heart rate estimation with less than 1.6% error, despite the model only locally approximating the true physical system.
This paper investigates the validity of utilizing Widrow's Active Noise Cancellation (ANC) in the context of motion artifact reduction for photoplethysmogram (PPG) sensors. The ANC approach has previously been applied to the PPG problem, but little consideration has been given to the validity of the ANC signal corruption assumptions and in what motion range the algorithm works. The ANC validity testing is done in the form of impact (approximate impulse) testing of the physical PPG system and comparing with the modeled response for a range of motion amplitudes. The testing reveals that the identified corruption model does not generally represent the true physical system, but locally approximates the true system. Testing shows that if a similar motion amplitude is used for model tuning as the impact test, an average peak deviation of 5.2% is obtained, but if motion amplitude that is smaller than the impact amplitude by a factor of 5, the peak deviation is 15%. Finally, after ANC filtering motion corrupted data, heart rate can be estimated with less than 1.6% error.
Wood et al. (Tue,) conducted a other in Motion artifacts in photoplethysmogram (PPG) sensors. Widrow's Active Noise Cancellation (ANC) was evaluated on Heart rate estimation error after ANC filtering. Active Noise Cancellation filtering of motion-corrupted PPG data enabled heart rate estimation with less than 1.6% error, despite the model only locally approximating the true physical system.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: