Coded Excitation of Guided Ultrasonic Waves in Long Bone for Assessment of Fracture Depth

Abstract

Single pulse (SP) excitation is the conventional method in long bone assessment which uses ultrasonic guided waves (GWs). However, the SP excitation will result in a small amplitude and low signal-to-noise ratio (SNR) of received signals due to the high attenuation during propagation. To solve the problem, this paper used coded excitation instead, which could help to improve the SNR of received signals. In simulations and in vitro experiments, the GWs were excited by 13 bits Barker code (BC) and 17 bits optimal binary code (OBC), then the received signals were decoded with weighted match filter (WMF) and finite impulse response least squares inverse filter (FIR-LSIF), respectively. The results showed good consistency between simulations and in vitro experiments. Besides ensuring the accuracy and authenticity of signals, the 13-bits BC and 17-bits OBC both got a larger amplitude and a better SNR than single pulse excitation. In the case of strong noise (SNR = 0 dB), the correlation coefficient between received signals and reference signals dropped to 0.70 when using single pulse excitation, while the correlation coefficient still remained above 0.95 for coded excitation. Furthermore, coded excitation modulated by 0.2 MHz sine wave had produced only two modes, L(0,1) and L(0,3), in received signals. When the depth of fracture varying from 0 to 0.5 mm, the energy mainly concentrated in L(0,3) mode. However, the energy mainly concentrated in L(0,1) mode when the depth of fracture varying from 0.5 to 5.0 mm. Therefore, the depth of fracture in long bone could be estimated based on energy transmission percentages.