Robust heart beat detection from photoplethysmography interlaced with motion artifacts based on Empirical Mode Decomposition

Many vital physiological features are embedded in photoplethysmography (PPG). Among them, heart beat carries the most significant importance for physiological monitoring in both the clinical and mobile health-care settings. However, motion artifact induced by finger and arm movement can corrupt the PPG signal significantly and cause serious false recognition of physiological features, leading to erroneous medical decision. In this paper, we propose a signal processing method based on multi-scale data analysis using Empirical Mode Decomposition (EMD) for the purpose of accurate heart rate extraction. Experiments with signals from Physionet database and the signals collected in our lab showed that our method can improve the accuracy of heart beat detection with period recovery rate at 84.68%.