The use of photoplethysmography (PPG) on the wrist to measure physiological indicators has attracted wide attention because of the portability and real-time characteristic of this technology. However, accurate estimation of the heart rate (HR) is difficult to realize using PPG because of the interference of motion artifacts. To address this problem, a method combining multichannel PPG signals is proposed. By using a peak selection method that combines several factors based on scores, the appropriate frequency is selected from the spectrum of the PPG signals. The chosen frequency is then considered as the HR. The approach exhibits high accuracy and speed. Experimental results for 12 training sets showed that with the proposed method, an average absolute error of 1.16 beats per minute (BPM) (standard deviation: 1.56 BPM) was obtained. Therefore, the proposed approach is reliable for HR monitoring from PPG during high-intensity physical activities. It can be applied to smart wearable devices for fitness tracking and health information tracking.
Keywords: Adapt filter; HR; Motion artifacts; Peak selection; Photoplethysmography (PPG).