Trans-rectal thermo-acoustic computed tomography: An initial in silico study

Background: The purpose of this in silico study is to demonstrate thermo-acoustic computed tomography (CT) based reconstruction of frequency-dependent true electrical conductivity distribution in a trans-rectal axialimaging geometry. Since cancerous tissue is expected to exhibit different conductivity profile compared to normal tissue, reconstructing conductivity based on thermoacoustic CT in a trans-rectal geometry has a potential for prostate cancer detection.Methodology: A trans-rectal axial-imaging geometry is illuminated by an electromagnetic (EM) point source at a microwave frequency. The source is located on a transrectal EM applicator close to the rectal wall. The applicator also houses a convex-array of point acoustic receivers that capture the acoustic pressure generated within the geometry as a result of EM illumination. The finite element method (FEM) along with an absorbing boundary condition is applied for solving the electric field (E-field) distribution, the power loss density and the acoustic pressure. The Levenberg-Marquardt regularization scheme is applied to reconstruct the conductivity distribution by decoupling the E-field from the power loss density.Results: For an excitation frequency of 915 MHz, various 2-D reconstructed images based on a 2:1 conductivity ratio between the background and object in a trans-rectal geometry of 40 mm radius are shown. Both single and double objects of 3 mm radius positioned at 4, 7, 10 and 15 mm depth with respect to the acoustic receiver are considered. The quality of the reconstructed image is shown to be object-depth dependent. The effect of different levels of Gaussian noise on the reconstructed images is shown. The contrast-to-noise ratios (CNRs) of the reconstructed images for the objects of different sizes and depths are also computed.Conclusions: Feasibility of recovering heterogeneous conductivity distribution in a trans-rectal axial-imaging geometry by thermo-acoustic CT is demonstrated in silico. The results implicate an alternative imaging mechanism for prostate cancer detection.