Inverse solution regularized with the edge-preserving constrain for NIR DOT

A well known problem associated with the diffuse optical tomography (DOT) is the reconstructed optical-property images suffer from low spatial resolution due to the diffusive nature of the light. It is noted that image reconstruction to find a inverse solution for DOT is an ill-posed problem [1]. Tikhonov regularization is a method of incorporating a priori assumptions or constraints about the desired solution, which converts an ill-posed problem into a well-posed problem. An iterative solution to the optimization problem was developed using the Tikhonov regularization with the optimal constraints regarding as a prior knowledge into the objective function. In this study, we focused on the edge-preserving constraint whereas concerning the improvement on the spatial resolution, rather than using structural prior information as the constraint in the objective function [2-6].