Quantitative non-linear ultrasonic imaging of targets with high acoustic impedance contrast—Application to bone imaging

Abstract

This study focuses on the ultrasonic imaging of high impedance acoustical contrast targets. The aim is to obtain information about shape, dimensions, and sound speed profile of the studied objects. One domain of application is the characterization of long bones. Quantitative information about the acoustic properties of bones tissues are of great interest for diagnosing or treatment monitoring of bone diseases. Inverse scattering problems of this kind are non-linear and various approximations can be used to linearize the scattering equations. Classical methods based on the first-order Born approximation give good results for weakly scattering targets but fail when it comes to give a quantitative information especially for high impedance contrast targets such as bones. In the inversion algorithm proposed here, Green's theorem is used to obtain a domain integral representation of the scattered field. An iterative non-linear algorithm minimizing the discrepancy between the measured and computed scattered fields is used to reconstruct the sound speed profile in the region of interest. The minimization process is performed using a conjugated-gradient method. An experimental study was performed with targets made of paraffin and with lamb bones. Images of the sound speed profile obtained by inversion of experimental data are presented and discussed in both cases.