#TITLE_ALTERNATIVE#

• Main Author: CAHYADI (NIM 23805001), HARSONO
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/10558
• Institution: Institut Teknologi Bandung

Diffuse Optical Tomography (DOT) refers to the imaging technique which employs near infra-red light radiation as the illuminator to produce input signal. DOT based on physical phenomena of light-medium interaction which is governed by two optical parameter of the medium, those are scattering and absorption coefficients.<p> <br /> <br /> <br /> <br /> <br /> <br /> <br /> One of reconstruction methods for DOT is Model Based Iterative Image Reconstruction (MoBIIR). It generally consists of three major problems : forward model, comparator, and optimator. Forward model performs numerical approach using the Finite Element Method (FEM) to provide predicted data. Comparator calculates the difference between measured data and predicted data using squared error method and it states as objectif function. Optimator performs optimization by minimizing an objectif function using Newton-Raphson method and then updates model parameters according to the result.<p> <br /> <br /> <br /> <br /> <br /> <br /> <br /> There are some numerical problems in image reconstruction, especially in optimization process, due to unsuitable object model and limited number of measured data. This results in ill-posed condition to system matrix. To solve this problems, we employ a regularization technique.<p> <br /> <br /> <br /> <br /> <br /> <br /> <br /> The reconstruction algorithm is investigated due to some object : object with 56 elements and 37 nodes, object with 136 elements and 81 nodes, object with prior information and value limit of parameter distribution, and object with varying regularization constant as well.<p> <br /> <br /> <br /> <br /> <br /> <br /> <br /> In this study, we observe DOT for circular object, steady state domain system, and time independent case using numerical simulation. This study results in : first, the ill-posed problem in DOT image reconstruction can be solved by employing a regularization technique; second, objective function is a criteria to determine the iteration of DOT image reconstruction; third, the regularization constant in reconstruction algorithm may be differed due to the object.