Image-guided radiation therapy (IGRT) provides accurate radiation delivery to the tumor and is being commonly used for various cancer sites. The last 10-15 years have witnessed an explosion of techniques for image guidance. On-board x-ray imaging evolves from two-dimensional to three-dimensional, and cone-beam computed tomography (CBCT) is being increasingly implemented on radiation therapy machines. Nonetheless, the current use of CBCT is greatly hindered by its poor image quality mainly due to scatter artifacts, and the clinical applications are limited to treatment setup. Quantitative CBCT images with high HU accuracy, which are particularly important for dose verification and adaptive radiation therapy, are still not achievable. The x-ray scatter contamination increases as the size of the irradiated volume increases. The large scatter signals in cone-beam projections lead to a contrast reduction and severe shading/streak artifacts in the CT images, which is considered the fundamental limitation of CBCT image quality. Scatter correction has been a very active research area in the last twenty years. The problem is challenging in human imaging since the scatter- to-primary ratio (SPR) is large and the imaged object has high heterogeneity. A standard solution still remains unclear. To facilitate the applications of CBCT in radiation therapy, we propose a new scatter correction method based on a unique feature of radiation therapy. With much smaller inherent scatter signals, diagnostic multi-detector CT (MDCT) obtains accurate CT images and is routinely used for treatment planning. Using the MDCT images as the "free" prior information of the patient, we estimate and correct for scatter in CBCT projections using simple and efficient image processing techniques. Our approach is distinct from other CBCT calibration methods, which totally rely on the MDCT image for providing the patient anatomic details. We use limited patient information from the MDCT image, and the merit of CBCT-based treatment monitoring is therefore retained. Our preliminary phantom study on a well-characterized tabletop system has shown that the proposed scatter correction reduces the CT number error from ~350 HU to ~5 HU and substantially increases the visibility of low-contrast objects. The overall goal of this research program is to extend and optimize our approach on the tabletop system, and to demonstrate that we can achieve a CT number accuracy of 20 HU on patients on a clinical CBCT system and use the improved CBCT for accurate tumor delineation and dose calculation. Our method has a potential to be a disruptive technology that removes hurdles toward high- performance IGRT and accurate dose monitoring for adaptive radiation therapy.

Public Health Relevance

The project aims to provide a correction solution to one of the fundamental physical processes, x-ray scatter, which limits the quality of cone-beam Computed Tomography (CBCT) imaging and its applications of image guidance in the current radiation therapy. If our method proves to be effective, it can remove the shading and streak artifacts in a CBCT image and make small and low-contrast tumors more visible. This project is very significant in that it could lead to a breakthrough toward accurate monitoring of delivered radiation dose on the tumor, which makes implementations of many new radiation therapy techniques possible.

National Institute of Health (NIH)
National Institute of Biomedical Imaging and Bioengineering (NIBIB)
Exploratory/Developmental Grants (R21)
Project #
Application #
Study Section
Biomedical Imaging Technology Study Section (BMIT)
Program Officer
Lopez, Hector
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Georgia Institute of Technology
Engineering (All Types)
Schools of Engineering
United States
Zip Code
Zhu, Lei; Niu, Tianye; Petrongolo, Michael (2014) Iterative CT reconstruction via minimizing adaptively reweighted total variation. J Xray Sci Technol 22:227-40
Niu, Tianye; Ye, Xiaojing; Fruhauf, Quentin et al. (2014) Accelerated barrier optimization compressed sensing (ABOCS) for CT reconstruction with improved convergence. Phys Med Biol 59:1801-14
Niu, Tianye; Al-Basheer, Ahmad; Zhu, Lei (2012) Quantitative cone-beam CT imaging in radiation therapy using planning CT as a prior: first patient studies. Med Phys 39:1991-2000
Zhu, Lei; Zhang, Wei; Elnatan, Daniel et al. (2012) Faster STORM using compressed sensing. Nat Methods 9:721-3
Niu, Tianye; Zhu, Lei (2011) Scatter correction for full-fan volumetric CT using a stationary beam blocker in a single full scan. Med Phys 38:6027-38