Nonlinear inversion of ultrasonic guided waves for in vivo evaluation of cortical bone properties

Abstract

Ultrasonic guided waves (UGWs), which propagate throughout the entire thickness of cortical bone, are attractive for the early diagnosis of osteoporosis. However, this is challenging due to the impact of soft tissue and the inherent difficulties related to multiparametric inversion of cortical bone quality factors, such as cortical thickness and bulk wave velocity. Therefore, in this research, a UGW-based multi-parameter inversion algorithm is developed to predict strength-related factors. In simulation, a free plate (cortical bone) and a bilayer plate (soft tissue and cortical bone) are used to validate the proposed method. The inversed cortical thickness (CTh), longitudinal velocity (V L) and transverse velocity (V T) are in accordance with the true values. Then four bovine cortical bone plates were used in in vitro experiments. Compared with the reference values, the relative errors for cortical thickness were 3.96%, 0.83%, 2.87%, and 4.25%, respectively. In the in vivo measurements, UGWs are collected from the tibias of 10 volunteers. The theoretical dispersion curves depicted by the estimated parameters (V T, V L, CTh) match well with the extracted experimental ones. In comparison with dual-energy x-ray absorptiometry, our results show that the estimated transverse velocity and cortical thickness are highly sensitive to osteoporosis. Therefore, these two parameters (CTh and V T) of long bones have potential to be used for diagnosis of bone status in clinical applications.