Liver cancers, both primary and metastatic, are increasing in incidence and are associated with significant mortality. Stereotactic Body Radiotherapy (SBRT) has been established as an effective, safe, and feasible first- line option in the local control of unresectable hepatic malignancies. Nonetheless, a large margin (typically ~1cm) has to be used in current liver SBRT to accommodate tumor positioning uncertainty under cone-beam CT (CBCT) image guidance. Because of respiratory motion and vanishing tumor contrast, the tumor target cannot be visualized in CBCT. Inferior tumor positioning approach based on anatomical or implanted surrogates are clinical standard, leading to substantial tumor position uncertainty and a typical margin size of ~1cm. Consequently, high dose to a large volume of normal tissue is delivered, causing a toxicity concern, especially in patients with liver dysfunction caused by cancer and/or treatments is more substantial. In addition, normal tissue toxicity limits further dose escalation to improve clinical benefits. This issue is expected to become more severe, when extending SBRT to a wider patient population, e.g. those with a large tumor size. Several emerging imaging approaches have showed potential to improve image guidance accuracy but also encountered challenges. To date, there is no approach that can provide accurate, reliable, and clinically translatable image guidance for liver SBRT. Recently, our group has made a breakthrough towards reconstructing elemental composition image using a standard CBCT platform. Employing a kVp-switching technique, a novel image reconstruction method with spatial and spectral image regularization, as well as a sparse-dictionary based element decomposition method, we achieved ~3% accuracy in elemental composition as tested in phantom studies. We have also accumulated extensive experience in reconstructing high-quality CBCT images under respiratory motion. Armed with these successes, the overall goal of this study is to develop a novel element-resolved and motion-compensated (ERMC-) CBCT to image iodine contrast agent using only 20% contrast injection in a standard treatment planning CT scan for precise (uncertainty <2mm) image guidance in liver SBRT. We will pursue three specific aims (SAs): SA1. Develop the overall ERMC- CBCT system. SA2. Optimize scan parameters via phantom studies. SA3. Perform studies in 10 patient cases to test safety, feasibility, and tumor positioning accuracy of ERMC-CBCT based image guidance. The innovation of this project is a novel ERMC-CBCT system and its application for a clinically significant problem of tumor localization in liver SBRT. Besides the significance of substantially improved localization accuracy and therefore clinical potential of normal tissue sparing and dose escalation, our project also holds the significance of utilizing CBCT to its maximal potential for many other advanced image guidance tasks and quantitative applications. The ERMC-CBCT system is developed on a conventional CBCT platform, the most widely available image-guidance platform in radiotherapy, ensuring its translatability.

Public Health Relevance

Stereotactic body radiotherapy (SBRT) has been established as an effective, safe, and feasible first-line option in the local control of unresectable hepatic malignancies. A large margin of ~1cm has to be used in current liver SBRT to accommodate tumor positioning uncertainty under cone-beam CT (CBCT) image guidance. This study will develop a novel element resolved motion compensated CBCT system to image low-dose iodine contrast agent. Successful completion of this study will lead to a safe, clinically feasible and accurate image guidance approach that substantially reduces the tumor positioning uncertainty, therefore reducing margin size and normal tissue toxicity.! !

National Institute of Health (NIH)
National Cancer Institute (NCI)
Research Project (R01)
Project #
Application #
Study Section
Biomedical Imaging Technology Study Section (BMIT)
Program Officer
Farahani, Keyvan
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Texas Sw Medical Center Dallas
Schools of Medicine
United States
Zip Code
Shen, Chenyang; Gonzalez, Yesenia; Chen, Liyuan et al. (2018) Intelligent Parameter Tuning in Optimization-Based Iterative CT Reconstruction via Deep Reinforcement Learning. IEEE Trans Med Imaging 37:1430-1439