Effects of polydispersity and high scatterer concentration on quantitative ultrasound estimates
Résumé
Quantitative ultrasound (QUS) techniques for the estimation of tissue microstructures are based on the frequency analysis of backscattered signals from the tissue. One of these QUS methods relies on theoretical scattering models that match a theoretical backscatter coefficient (BSC) to the measured BSC in order to estimate QUS parameters (such as the effective scatterer size and acoustic concentration). The scattering models generally assume that the tissue under consideration contains identical scatterers in sparse medium and may provide QUS parameters that are not representative of the actual tissue microstructure when the medium is polydisperse in scatterer size and/or dense. Our goal is to study the effects of scatterer polydispersity on QUS estimates by considering three scattering models: the monodisperse sparse model, the polydisperse sparse model and the monodisperse concentrated structure factor model (SFM). In that aim, simulations of backscatter coefficients are conducted with different scatter size distributions for sparse or moderately dense media. The QUS parameters are estimated at different center frequencies f0 (from 10 MHz up to 50 MHz) by fitting the simulated backscatter coefficient in the frequency bandwidth [0.5f0 − 1.5f0], i.e. assuming a transducer with 100% usable bandwidth. The choice of scattering models have a high impact on the QUS parameter estimates as function of frequencies for values of ka less than 1.2.
Origine : Fichiers produits par l'(les) auteur(s)