Research On Image Reconstruction For Time-Domain Breast Diffuse Optical Tomography Based On Panel Detection And Finite-Difference Method

As the incidence and mortality of breast cancer grow, the research on realizing regular examination of breast has become one of the focused topics. The main object of this paper is to develop an algorithm using finite-difference method (FDM) based on the diffuse equation (DE) according to the panel-compressing detection mode, and to apply it to the diffuse optical image reconstruction. The DE is derived from a more accurate model of describing light propagation in turbid media– the radiative transport equation. FDM is an effective method of solving partial equation numerically, which is perfectly suitable for modeling the panel dictation mode. The result of the forward model is compared with analytical result to demonstrate its validity.Based on the above forward model and the optimization expression of the reconstruction, Newton-Raphson linear iteration scheme has been employed to solve the inverse problem. Due to the requirement of high spatial resolution by the clinic implementation, difficulties come up with the calculation of Jacobi matrix. An optimal perturbation algorithm based on variational principles and General Pulse Spectrum Technique (GPST) has been presented in this paper to avoid calculating Jacobi matrix directly. In this paper, the iterative row-action based algebra reconstruction technique (ART) is adopted as a regulation to take place of the direct solving of the ill-posed linear equation set.The reconstruction result of the numerical simulation under three-dimensional model with target of different optical coefficient and volume illustrates the validity and reliability of the reconstruction algorithm proposed in this paper. Due to the time limitation, only absorption coefficient related imaging reconstruction has been tested. Spatial resolution, grayscale resolution, as well as robustness are evaluated, which lead to a satisfactory conclusion that this algorithm is suitable for further application into diffuse optical tomography with the panel detection mode. But before applying to clinic implementation, more work should be done with further testifying and algorithm improvement.