Simulation of ultrasonic scattering in breast tissue based on cell and tissue morphology.

Abstract

High-frequency ultrasonic spectroscopy is a promising approach for the in vivo detection of cancer in applications such as margin assessments during breast cancer surgery. To advance this approach, a multiple scattering model was developed to more fully understand and analyze ultrasonic backscatter data in the 10–100-MHz region from breast tissues in various stages of tumor progression. The model uses an iterative multipole approach to simulate the multiple scattering of shear and longitudinal waves between nucleated cells in a histologically accurate tissue representation. Previously, concentric spheres were used to model the nucleus and cytoplasm of cells organized in complex, hierarchical tissue structures consisting of alveolar cavities and up to 2075 cells [Doyle et al., J. Acoust. Soc. Am. 125, 1751–1767 (2009)]. The simulations used addition theorems to model multiple scattering and showed significant spectral changes that correlated with tumorigenesis. To increase the fidelity of the simulations, the scattering model has been expanded to incorporate spheroidal and cuboidal cells, non-centered nuclei, mixed tissue types, and duct-like channels representing the three-dimensional luminal structure in breast tissue. Theory, simulation results, and experimental data will be presented. [Work supported by NIH 1R21CA131798-01A1.]