Extracting Heartbeat Intervals Using Self-adaptive Method Based on Ballistocardiography(BCG)

摘要

Ballistocardiogram (BCG) could reflect mechanical activity of cardiovascular system instead of ECG. And it is often acquired by sensitive mattress or chair without any constraints and limitations, but it contains many noise because of the impact of body and acquired equipment, those questions make heart rate detection difficult from the original BCG. In the paper, we propose an adaptive method which is used to extract heartbeat intervals (RR), and the method acquire automatically input parameters of Ensemble Empirical Mode Decomposition (EEMD) algorithm, and then decompose BCG signal using EEMD algorithm, and select adaptively decomposition component of BCG signal, whose periodicity is in accordance with the cardiac cycle completely as the target signal. Furthermore we detect the peak points and calculate the heartbeat intervals series using the target signal. In the result, the proposed method is tested using the BCG datasets from 18 subjects, including 8 females and 10 males (age 20-72). Finally, the heart rate from BCG will be compared with ECG, and the results are satisfactory and have a high accuracy.