Session P73.4

Estimation of Heart Rate Variability with Ballistocardiogram

JH Shin*, KH Hong, KK Kim, KS Park

Seoul National University
Seoul, Korea

In this study the dynamics of heart was analyzed with ballistocardiogram (BCG) by non-invasive cardiopulmonary activity measurement method. We investigated whether this heart activity signal is reflected in the surface electrocardiogram (ECG) and whether additional information can be obtained from BCG.
To get reliable heart activity parameters in BCG, we designed a detection algorithm for finding J peak which corresponding with R peak of ECG. The algorithm was based on dilation window method to absorb low-frequency noise components and to enhance smoothing effect and adaptive threshold technique to reject noise peaks. The detected J peak intervals were computed for evaluating the RR rhythm and the heart rate variability (HRV) of BCG was analyzed in frequency domain.
The performance of J peak detection algorithm was evaluated with five minute BCG and ECG data pairs which measured from a bed type cardiopulmonary activity measurement sensor and simultaneously acquired a conventional Ag/AgCl electrode when every ten normal adult lie in a bed. The detected J peaks were compared with R peak and paired t-test was carried out in order to analyze the HRV parameters, LF, HF and LF/HF ratio. The results show a highly corresponding performance of J peak detection algorithm with ECG R peak and the correlation coefficients of HRV were reliable.

(Abstract Control Number: 175)