Vital Sign Signal Extraction Method Based on Permutation Entropy and EEMD Algorithm for Ultra-Wideband Radar

Degui Yang,Zhengliang Zhu,Buge Liang

doi:10.1109/access.2019.2958600

Abstract

The noncontact measurement of vital sign signals is useful for medical care, rescuing disaster survivors from ruins and public safety. In this paper, a novel vital sign signal extraction method based on permutation entropy (PE) and ensemble empirical mode decomposition (EEMD) algorithm is proposed. The proposed algorithm analyzes the permutation entropy of radar-received pulses; the range between a human target and ultra-wideband (UWB) radar can be obtained by permutation entropy. Permutation entropy represents the complexity of signals, so we can use PE to select and recombine human life signals that are distributed in the adjacent distance gate. Moreover, EEMD algorithm is adopted to decompose the combined signal into intrinsic mode functions (IMF), and both the respiration and the heartbeat signals are reconstructed by IMF via reaching the energy threshold in the time domain. Experiments are carried out using UWB radar. Compared with traditional algorithms, the proposed algorithm can be used to extract the range and frequency information of human targets efficiently and accurately.

Highlights

In recent years, the needs of human noncontact measurement of vital signs by using radar is increasing; relevant research results are applied to the medical care of patients [1], [2], the rescue of disaster survivors from ruins [3], [4], public safety [5]–[7], and so on
We propose a vital sign signal extraction method based on permutation entropy (PE) and an ensemble empirical mode decomposition (EEMD) algorithm for UWB radar
The signals are sampled by the echo acquisition module, and converted into digital signal by an analog-to-digital converter (ADC) and stored by a field programmable gate array (FPGA)

Summary

INTRODUCTION

The needs of human noncontact measurement of vital signs by using radar is increasing; relevant research results are applied to the medical care of patients [1], [2], the rescue of disaster survivors from ruins [3], [4], public safety [5]–[7], and so on. D. Yang et al.: Vital Sign Signal Extraction Method Based on Permutation Entropy and EEMD Algorithm for Ultra-Wideband Radar cumulants of reconstructed signals were calculated to improve measurement accuracy [16]. In traditional algorithms for human vital sign extraction, the respiratory and heartbeat frequency are obtained by Fourier spectrum analysis of the single-frame signal. We need long-term observation data; this need, reduces the efficiency of radar To solve this problem, we propose a vital sign signal extraction method based on permutation entropy (PE) and an EEMD algorithm for UWB radar. By choosing and combining the signals on these distance gates based on PE values, more vital signs information can be obtained than single-frame signals, and the observation time can be reduced.

UWB RADAR VITAL SIGN SIGNAL MODEL

RANGE DETECTION AND ECHOES SELECTION

CONCLUSION