Artifacts in wearable photoplethysmographs during daily life motions and their reduction with least mean square based active noise cancellation method

Signal distortion of photoplethysmographs (PPGs) due to motion artifacts has been a limitation for developing real-time, wearable health monitoring devices. The artifacts in PPG signals are analyzed by comparing the frequency of the PPG with a reference pulse and daily life motions, including typing, writing, tapping, gesturing, walking, and running. Periodical motions in the range of pulse frequency, such as walking and running, cause motion artifacts. To reduce these artifacts in real-time devices, a least mean square based active noise cancellation method is applied to the accelerometer data. Experiments show that the proposed method recovers pulse from PPGs efficiently.