A comparison of three imaging algorithms on reconstructing tumor target in acoustic heterogeneous tissue with realistic numerical breast phantoms in microwave induced thermo-acoustic tomography

Accurately reconstruct the tumor target in an acoustic heterogeneous tissue is a hot research issue in thermo-acoustic tomography (TAT). In this paper, we present three imaging algorithms for microwave induced thermo-acoustic tomography (MITAT): time reversal mirror (TRM), active adjoint modeling (AAM) method and iterative reconstruction method (IRM). The performance of the three imaging methods for reconstructing tumor target in heterogeneous breast tissue are verified and compared by using the anatomically realistic numerical breast phantoms derived from T1-weighted MRIs of prone patients. Through the simulation results, the IRM proves to have better imaging results than TRM and consume much less computing time than AAM. The IRM is a promising approach for imaging tumor target in heterogeneous breast tissue in TAT.