Determining breast pathology in surgical margins with high-frequency ultrasound: phantom and numerical simulations

Abstract

Two parameters in high-frequency ultrasound (20-80 MHz) have been found to be sensitive to a range of pathologies in resected margins from breast conservation surgery: The number of peaks (the peak density) in the waveform spectrum and the slope of the Fourier transform of the waveform spectrum. Previous studies have indicated that peak density and slope may correlate to microscopic heterogeneity in tissue structure, which is modified by atypical and malignant processes. To test this hypothesis, through-transmission and pulse-echo measurements were acquired from gelatin-based phantoms containing polyethylene microspheres and nylon fibers (2.5-10% volume concentration). Multipole methods were also used to model through-transmission measurements of tumor progression in lobular carcinoma in situ. The simulated breast tissue contained 1000-2000 nucleated cells with random lobular cavities. The peak densities of the heterogeneous phantoms were significantly greater than those of the homogeneous control samples, whereas the slopes were less. Similarly, the models produced spectra with peak densities that increased with malignant cell proliferation. The results are consistent with breast tissue data, and provide a physical mechanism for the use of peak density and slope in the imaging of breast tissues with atypical and malignant pathologies. This work was supported by Utah Valley University.