An Experimental Study of Diffraction Tomography under the Born Approximation

Abstract

Diffraction tomography under the Born approximation was studied experimentally to help assess its potential utility for clinical imaging, particularly in the detection and classification of breast disease. The Born approximation is convenient to use since, being linear, it leads to simple and efficient inversion algorithms. However, its appropriateness for clinical imaging is questionable. In these experiments, particular care was taken to minimise errors of the data acquisition and image reconstruction processes in an effort to isolate the effects of the choice of the Born scattering model on the overall performance of diffraction tomography. Diffraction tomograms of a tissue mimicking breast phantom are compared with both X-ray and ultrasound computed tomograms The results indicate that, even under conditions when the assumptions of the Born approximation are violated, useful images can be obtained. Such images though not quantitatively accurate maps of the complex refractive index, allow identification of the major internal structures.