Demonstration and optimization of X-ray StaticCodeCT
Date
2022
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Since Sir Godfrey Hounsfield invented the first CT scanner in 1967, the practice of X-ray computed tomography (CT) has exploded. X-ray CT has become widely employed in medical diagnosis in the last half-century. As more information about X-ray radiation becomes available, so does concern about the potential for radiation exposure to cause cancer. To alleviate some of this problem, state-of-the-art techniques have focused on hardware design constraints and reconstruction algorithm development. To block the radiation, coded aperture X-ray imaging, for example, inserts coded aperture masks in front of the X-ray source at each view angle. The image is then reconstructed by solving a l1-based optimization problem, due to the inverse problem's ill-posedness. Unfortunately, with this approach, the best reconstructions are obtained when distinct masks are employed in each view angle, which makes it impractical in real-world applications. In this thesis, a single-static coded aperture is used in StaticCodeCT, a newly developed low-dose tomography design. It utilizes data with high correlation in the projection range and algorithms developed to accurately approximate the detector's missing measurements across the spatial and axial dimensions. Using iterative image reconstruction or so called the FDK algorithm, the object from a complete set of cone-beam synthesized projections is successfully recovered. The effect of StaticCodeCT system was investigated before by simulating the coded aperture mask on CT data from experiments. A preliminary cone-beam X-ray CT system with a coded aperture holder is used in this study to offer test bench reconstructions. We looked at the quality of the reconstruction by using coded apertures made with two different techniques: metal additive manufacturing and cold-cast manufacturing with plastic mold. Additionally, based on the measurement estimation methodology, we offer an optimization strategy for designing the coded aperture pattern. The observed results suggest that real-world deployment of StaticCodeCT systems is possible.