Study On The Optimization Inversion Algorithm Of Photon Density Wave Scattering

Diffuse optical imaging is an emerging technique for functional imaging of biological tissue. If there exists an abnormity (e.g. tumor) in a bio-tissue, measurements of visible or near-infrared light scattered from the tissue will be different. Therefore, the measurements can be used to detect and localize the abnormity by solving the inverse problem according to photon density wave scattering model. For the goal of solving the inverse problem like this, it is necessary to develop an effective way to describe the behavior of light propagation in the tissue (highly scattering media) and to design a proper algorithm to reconstruct the structure of the tissues.The simulation results of the analysis resolution are shown based on the presented typical optical parameters of human being's forearm tissue. The results validate the theory of the diffuse photon density wave. More importantly, the site, the size and the absorption coefficient of the abnormity have a great influence on the photon flux, which is very useful to the solution of the inverse problem.The inverse problem can be solved by minimizing the difference between the measured data and data which are simulated using the scattering model, then the distribution of optical parameters can be obtained. The inverse problem above is actually an optimization problem. Therefore, optimization models are devised for different cases by use of the analytic expression of the photon flux and the least squares (LS) method.Theoretically, many optimization methods can be employed to solve the optimization problems. However, in consideration of the limit of computer storage space and the complexity of the problem, two requirements must be satisfied: 1) to reduce the calculations and storage for the derivatives of the forward operator; 2) to reduce solving times of the forward problem. For this purpose, a modified Harmony Search (HS) is presented and employed to seek the possible solutions. The constraint conditions can be conveniently involved into the algorithm. And the algorithm can be easily implemented and does not need any linear search procedure, which can reduce procedure for solutions of the forward problems. The proposed modified HS can be used for solving the inverse problems with higher efficiency and accuracy compared with the conventional HS in which the parameters such as the harmony memory considering rate (HMCR) and the pitch adjusting rate (PAR) are difficult to determined.Based on the simulated measuring data obtained from solving the forward problem, the numerical experiments for comparisons of the conventional HS and the modified HS are given. The results show that the improved algorithm can be used to detect the abnormity embedded in the tissue with a good agreement with the presupposed parameters. It can be concluded that the improved HS algorithm is more effective than the conventional HS for the inverse problems similar to the detection using photon density wave.The research consequences may offer a theoretical and experimental reference to the bio-tissue imaging and clinical diagnose using diffuse photon density wave. It can also be expected that the improved HS with adaptive parameter selection can be applied to other optimization problems in various fields.