Compressed sensing in microwave induced thermo-acoustic tomography

Aiming to reliably reconstruct an image with fewer measurements, we adapt compressed sensing (CS) in microwave induced thermo-acoustic tomography (MITAT). Most of current existing reconstruction algorithms for MITAT employ circular detection and require a great deal of measurements, which imply long acquisition times. In this paper, CS is implemented to reconstruct an image of electromagnetic absorption distribution with fewer numbers of detection angles and Fourier samples by using gradient projection for sparse reconstruction (GPSR) algorithm. Some numerical simulation experiments are given to demonstrate the excellent performance of CS in MITAT.