Comparison of two image reconstruction algorithms for microwave tomography

Authors


Abstract

[1] Two image reconstruction algorithms for microwave tomography are compared and contrasted. One is a general, gradient-based minimization algorithm. The other is the chirp pulse microwave computed tomography (CP-MCT) method, which is a highly computationally efficient reconstruction method but also a method best suited for low contrasts. The results of the simulations show that when imaging high-contrast objects, such as a breast cancer tumor, reconstructions made are comparable to results from the minimization algorithm below a contrast of about 10%. The simulations, however, show that the reconstructions made by the CP-MCT method are very robust to noise. The reconstruction of the conductivity using the minimization algorithm, on the other hand, is very sensitive to the level of noise. In spite of a strong degradation in the conductivity reconstructions, the corresponding permittivity reconstructions do not show the same sensitivity to the noise level.

Ancillary