Diffraction Tomography Breast Imaging System: Patient Image Reconstruction and Analysis

Conclusions

Analysis of patient breast images from a clinical trials series of 25 patients suggest that ultrasound diffraction tomography has the potential for discriminating between different tissue types. On the basis of very limited data, it further appears that image contrast values of malignant lesions provide sufficient discrimination from normal breast tissue to be potentially of diagnostic value. An analysis of tissue scatter amplitude distributions shows some correlation between the nature of the distributions, characterized by a Tissue Index, and predominant breast tissue classifications. Since this is reflective of the strength of scattering, the main utility of this analysis is potentially as an a priori tool for the selection of optimal reconstruction algorithms. The data are not sufficiently specific nor interpretable to be of diagnostic value.

The observed strong scattering of breast tissue and the object dimensions lead to phase shifts that are well beyond the validity conditions for the Born approximation. We conclude that our hybrid reconstruction techniques extend the linear range of the propagation constant that can be reconstructed but lose the ability to separate it into components of speed of sound and attenuation.