Search

Human breast is a heterogeneous medium for microwave signal. Breast cancer detection using microwave imaging is done based on signal scattered by breast tissues at different frequencies. Wave propagation direction is extremely important in heterogeneous medium like human breast. In this paper, the effect of wave propagation direction on the dielectric profile reconstruction is simulated in the presence of noise. x and Y directed transverse electric (TE) waves are considered for numerical breast phantom heterogeneity exploitation. Wave propagating in Y direction results into better dielectric profile reconstruction than X directed wave. Signal to noise ratio is very crucial for microwave imaging because information resides in low power scattered electric signal. Results show that SNR of at least 30 dB is required to detect cancer by solving extremely under-determined system of scattering equations. [ABSTRACT FROM AUTHOR]

In microwave imaging, accuracy of breast cancer detection depends on complex permittivity profile reconstruction in breast. Inverse scattering problem is solved to reconstruct complex permittivity profile of breast. In this paper, computation time to solve inverse scattering problem is reduced by exploiting symmetry present in breast models using group theory. Forward problem is solved using method of moments. Levenberg-Marquardt algorithm is used to solve inverse scattering problem with and without group theory. Results show that computation time is reduced considerably by exploiting symmetry present in breast models using group theory. At higher SNR, error in complex permittivity reconstruction with group theory is approximately same as error without group theory. [ABSTRACT FROM AUTHOR]