Akademik Veri Yönetim Sistemi
OPTICAL AND QUANTUM ELECTRONICS, cilt.50, sa.3, 2018 (SCI-Expanded)
In this work, new image reconstruction schema has been proposed for back-reflected diffuse optical tomography (DOT) geometry. 100 source and 100 detector points have been selected as bifurcated probe positions in xy plane. An x, y, and z cubic coordinate grid system has 10x10x30 mesh grids. In this work, a dynamic mesh grid concept has been introduced first. Each time source and detector positions are changed, x and y coordinate positions have been reassigned automatically. Center between source and detector positions has been recalculated in xy plane. x and y grid positions have been reassigned around this center point. When using predefined static mesh grid photon fluencies which are coming from Monte Carlo (MC) simulation output or diffusion equation are transferred into static voxels which is only partially correct. Hence there is no symmetry around source-detector matchup center in xy plane and so is not totally correct. In this work dynamic mesh grids have been created and photon fluencies have been assigned into automatically redefined dynamic mesh voxels. One dimensional depth profile has been tested for simplicity. Median values of transport functions for each z depth grids have been used. Transport functions of photon fluencies from each source to detector positions have been calculated and each voxel value has been assigned in three dimensional dynamic mesh grid array. We innovated the use of dynamic mesh grid array instead of using static mesh grids which were being used previously, and now have assigned the photon fluencies into the dynamic mesh grid voxels. This also gave us an opportunity to use different transport functions for one dimensional image reconstruction schema. Before, we were using the sum of transport weight functions which are derived from each of the z depth layers. Now we have opportunity to use mean or median values also. Thus we are centering the dynamic mesh grid array by finding the center of each source and detector position successfully. This work gave us encouragement for building upcoming tomography devices.