An investigation on regularization methods for sound source reconstruction of magnetoacoustic tomography (MAT)

The inverse problem of Magnetoacoustic Tomography (MAT) is often ill-posed in the sense that noise in the data may give rise to significant errors in the process of reconstruction. It is critical and significant to develop a feasible and effective method to deal with this ill-posed problem. In this paper, we introduce a regularization method combining Tikhonov filter function, unbiased predictive risk estimator with gradient descent method for sound source reconstruction. Through simulation studies, we find that the proposed method performs well enough to reconstruct source and remove noise.