Akademik Veri Yönetim Sistemi
El-Cezerî Journal of Science and Engineering, cilt.2, sa.2, ss.40-46, 2015 (Scopus)
Continuous wave reflectance diffuse optical tomography (CWRDOT) systems incapable for accurate depth localization, since physical nature of light. We have used a new approach to calculate the unknown optical properties of breast like tissue by getting weight data from Monte Carlo (MC) simulation codes. In this work, we have developed a depth normalization algorithm (DNA), which normalizes weight matrix elements for each source and detector couplings for separate depth layers. Banana functions for source-detector pairs have been created by multiplying individual fluence rates of each source and detectors. Then, the weight coefficients of the voxels in volume have been used in inverse problem solution with the Tikhonov regularization method to obtain absorption and scattering coefficients for each voxel of tissue phantoms. Tikhonov regularization method has been used as an inverse problem solution to reconstruct the tissue phantom concentrations. It has been shown that a contrast-to-noise ratio (CNR) is 0.85 to 2 depends on the depth of the inclusion and the average position error over all the depths between 0 to 4 cm are 12.5 %.
Continuous wave reflectance diffuse optical tomography (CWRDOT) systems incapable for accurate depth localization, since physical nature of light. We have used a new approach to calculate the unknown optical properties of breast like tissue by getting weight data from Monte Carlo (MC) simulation codes. In this work, we have developed a depth normalization algorithm (DNA), which normalizes weight matrix elements for each source and detector couplings for separate depth layers. Banana functions for source-detector pairs have been created by multiplying individual fluence rates of each source and detectors. Then, the weight coefficients of the voxels in volume have been used in inverse problem solution with the Tikhonov regularization method to obtain absorption and scattering coefficients for each voxel of tissue phantoms. Tikhonov regularization method has been used as an inverse problem solution to reconstruct the tissue phantom concentrations. It has been shown that a contrast-to-noise ratio (CNR) is 0.85 to 2 depends on the depth of the inclusion and the average position error over all the depths between 0 to 4 cm are 12.5 %.