Weight matrix analysis for back reflection continuous wave diffuse optical tomography (CWDOT) systems: translational method

Abstract

Back reflection continuous wave diffuse optical tomography (CWDOT) systems use forward model photon fluence rate distributions for inverse problem solution to be able to reconstruct images. Forward model weight functions are also known as photon fluence rate distributions inside the homogenous media. The linearized equation systems are used for CWDOT imaging. To be able to recover the unknown delta absorption coefficients over background inside the homogenous media linearized equation system has to be solved. In this work the forward model weight matrix functions have been calculated according to the photon distribution of radiative transport equation in theoretical physics. After the forward model is set, the mathematical inverse problem solution methods have to be applied. In literature, the solution of linearized CWDOT equation systems are nonlinear methods but they have never been popular. The main goal of this study is to be able to show alternative weight matrix generation method. The weight matrix generation is very important for linearized CWDOT systems. The Monte Carlo (MC) simulations have been used to create the weight matrix by using the radiation transport problem solution via Rytov or Born approximation. The photon is sent into the tissue, the step distances of each photons are computed and the scattering angles of photons have to be known. It is calculated by occurrence probability for each voxel by the Henyey-Greenstein phase function. The generated MC simulation data is migrated from the simulation environment to the image reconstruction algorithm environment. The weight matrix is successively generated by explained translational method.