DIFFUSE OPTICAL TOMOGRAPHY: AN ADVANCED MULTISTAGE INVERSE METHOD AND AN OPTIMIZATION METHOD OF SOURCE AND DETECTOR ARRANGEMENTS

Abstract

Near-infrared diffuse optical tomography (DOT) has been proven in the last few decades to be a non-invasive and non-ionizing promising tool for functional imaging of soft tissue. The main obstacles to further the clinical application of DOT are its poor spatial resolution and its inherent difficulty of retrieving optical property distribution in tissue. This thesis focused on the improvement of the image quality of DOT via an advanced multistage inverse method of fast computation and high accuracy for 3D large volume imaging, and an optimization method of source and detector arrangements based on the Cramer-Rao lower bound (CRLB) method to enhance the capacity of noise resistance and reconstruction stability. Both proposed methods were demonstrated to be superior to conventional methods by thorough simulation. An optimized rotatable imaging probe based on the CRLB method was also implemented and integrated into a time-resolved DOT system for tissue phantom imaging.