Back reflected laser tomography

Abstract

An optical imaging model has been studied. Monte Carlo Modeling of Light Transport (MCMLT) has been run for the head model and photon fluencies have been generated. Multi-layered head model has seven different layers which are hair, epidermis, dermis, skull, cerebrospinal fluid, white matter, and gray matter with consecutive order from surface to deeper tissues. Back-reflected laser imaging geometry has been selected. Geometric multi source-detector positions are on the flat imaging surface. Source-detector matches are creating optode array on the imaging tissue surface. Each source-detector coupling is representing one single equation for diffuse optical tomography (DOT) system. In this work we had 64 sources and 64 detectors which makes 4096 equations. The equation system is building forward model. Forward model has independent equations which are equal to source-detector couplings. Our back-reflected continuous wave (CW) laser DOT model has been tested with simulation data by using the Tikhonov inverse problem solution algorithm. Forward model transport functions have been calculated. It has been built according to source-detector positions and grids. The Tikhonov inverse problem solution algorithm has been tested with different λ regularization parameters. The best parameter has been used for image reconstruction algorithm. One inclusion has been put into deeper voxels then image has been reconstructed. The result has been illustrated in xy, xz, and xyz coordinate grids with different angles. It has been reconstructed successfully with the correct position and concentration as well. This work gave an idea about how source-detector matches are important for back reflected DOT devices. As long as source-detector numbers are more than enough, reconstructed images are becoming more realistic as compared to the real inclusion location. This work encouraged us to design new laser tomography devices for near future. The algorithm has been tested with simulation data. Image reconstruction algorithms are important part of DOT devices. Progressive efforts are being made in academic research institutes and facilities. Algorithmic modeling is becoming very important and researchers are working to explore mathematical and philosophical ways to get better results. This work is presenting a new algorithmic approach to the laser imaging devices.