Simulation of a tissue backscatter coefficient measurement experiment using the finite element method

Abstract

Ultrasound has been shown to be an effective imaging modality for investigating tissue health. For example, the backscatter coefficient (BSC) has been shown to be a biomarker of tumour response to therapy but is limited in its application clinically due to the difficulty of accurately acquiring attenuation and diffraction corrections. To assess the variability in BSC assessment, we present a finite element tool simulating a singleelement planar reflector substitution method estimate of the BSC of a set of simulated phantoms. BSC estimates were computed with errors <5% of the theoretical value for a range of source apertures and over a set of phantoms with varying scatterer number densities to within 10%. In addition, the firstorder amplitude envelope statistics of backscattered waves were shown to be commensurate with Rayleigh scattering models. These results indicate that the tool is accurate in its replication of soft tissue-like scattering, suggesting that it could be an opportune testing ground for investigating sources of variability in BSC measurements and the testing of algorithms and hypotheses for interrogating the scattering behaviour of soft tissues.