Analysis of linearized inverse problems in ultrasound transmission imaging

The purpose of this paper is to analyze the linearized inverse problem during the iterativesolution process of the ill-posed nonlinear inverse problem of image reconstruction for ultra-sound transmission imaging. We show that the conjugate gradient applied to normal equation(CGNE) method gives more reliable solutions for linearized systems than Tikhonov regular-ization methods. The linearized systems are more sensitive when treated by CGNE than byTikhonov regularization methods. The Tikhonov regularization is less effective at the be-ginning of the outer-loop iteration, where the nonlinearity is dominating while the conjugategradient for the linearized system stops earlier. Only when the linear approximation is goodenough to describe the whole system, Tikhonov regularization can fully play its role and giveslightly better reconstruction results as compared to CGNE in a very noisy case.