Cong Zhao

Graduation Semester and Year

2015

Language

English

Document Type

Thesis

Degree Name

Master of Science in Physics

Department

Physics

First Advisor

Mingwu Jin

Abstract

Cone-beam computed tomography (CBCT) imaging is widely used in radiation therapy for treatment planning and image-guided therapy. More x-ray scatter signals are detected by the two-dimensional detector arrays in CBCT than by the one-dimensional detector arrays in fan-beam CT (FBCT). Therefore, scatter contamination significantly degrades the image quality in CBCT and must be corrected for satisfying image quality.One approach for scatter correction is to use lead blockers. In such a method, the signal in the blocked regions is assumed to be the scatter signal and can be used to estimate the scatter signal in the unblocked region. However, this assumption is often violated because of the penumbra effect and detector response effect. In this study, we first investigated a moving-blocker based method to simultaneously estimate the scatter signal and reconstruct the complete volume within the field of view (FOV) from a single CBCT scan. Then we proposed to use deconvolution algorithms to improve scatter estimation of the moving-blocker method. A slanted edge method was used to estimate the point spread function (PSF) of the flat panel detector. The parametric PSF was used in simulation studies to deconvolve the blurred projection image to remove the finite x-ray source and the detector response effect and to get cleaner scatter signal in the blocked regions.Our results showed that with laborious parameter tuning, the scatter-induced shading artifacts are substantially reduced in CBCT, and the CT number errors in the selected regions of interest are greatly reduced using the moving-blocker method. The detector response function can be reliably measured and calculated using the slanted edge method. The deconvolved projection image shows better quality, more accurate intensity profiles, and much reduced overall root mean square errors compared to the blurred image in the simulation study. Our experiment results further demonstrate that Wiener and Richardson-Lucy deconvolution methods can significantly improve the scatter estimation compared to the direct inverse filtering method. Finally, The PSF derived from the slanted edge experiment does not take the penumbra effect into account, and the air-scan data from the CBCT projection images of moving blockers fail to derive a reliable PSF. An experiment to determine the combined penumbra and detector response effect and the ultimate effectiveness of deconvolution methods on scatter correction of CBCT reconstruction will be investigated in future work.

Disciplines

Physical Sciences and Mathematics | Physics

Comments

Degree granted by The University of Texas at Arlington

Download

Included in

Physics Commons

Share

COinS