Heartbeat and respiration monitoring based on FBG sensor network

Abstract

A fiber Bragg grating (FBG)-based acquisition method for heartbeat and respiration was designed for portable home-based cardiopulmonary physiological parameter detection. The ballistocardiogram (BCG) is susceptible to noise interference during the acquisition process. In order to improve the accuracy of heart rate and respiration rate solving, the adaptive K-value-based variational modal decomposition (K-VMD) algorithm is proposed. First, the algorithm obtains the optimal solution by calculating the difference frequency of the last component of the neighboring K-values. Second, a threshold value is set according to the frequency variation. Again, the optimal K value is automatically selected according to the threshold. The problem of over-decomposition or under-decomposition can be solved by using the optimal K-value decomposition. Finally, heart rate and respiration rate are estimated using the power spectrum based on Welch's method. In 15 comparison experiments, the K-VMD algorithm was in perfect agreement with the ECG signal 13 times, and the other three conventional methods were 9, 7, and 5 times, respectively. The maximum error of the K-VMD algorithm was 1 bpm, and the other three conventional methods were 2 bpm, 3 bpm, and 3 bpm, respectively. The maximum error in the respiratory rate detection of this algorithm in the position of the diaphragm is ±0.2 bpm. It can be seen that the present algorithm can effectively reduce noise interference and improve monitoring accuracy. And it can provide data support for obtaining the characteristic waveform of cardiac shock signals and respiratory signals. It has certain practical applications.