Encrypted login | home

Program Information

A Scatter Elimination Scheme for Cone Beam CT Using An Oscillating Narrow Beam

H Yan

H Yan1*, Y Xu1,2 , M Folkerts1 , X Jia1 , S Jiang1 , (1) UT Southwestern Medical Center, Dallas, TX, (2) Southern Medical University, Guangzhou, China


TH-A-18C-6 Thursday 7:30AM - 9:30AM Room: 18C

Purpose:While cone beam CT (CBCT) has been widely used in image guided radiation therapy, its low image quality, primarily caused by scattered x-rays, hinders advanced clinical applications, e.g., CBCT based on-line adaptive re-planning. We propose in this abstract a new scheme called oscillating narrow beam CBCT (ONB-CBCT) to eliminate scatter signals.

Methods:ONB-CBCT consists of two major components. 1) Oscillating narrow beam (ONB) scan and 2) partitioned flat panel containing multiple individual detector strips and their own readouts. Both the beam oscillation and detector partition are along the superior-inferior (SI) direction. During data acquisition, at a given projection, the narrow beam sweep through the detector region, and different portions of the detector acquires projection data in synchrony with the narrow beam. ONB can be generated by a rotating slit collimator design with conventional tube with single focal spot, or by directly using a new source with multiple focal spots. A proof-of-principle study via Monte Carlo simulation is conducted to demonstrate the feasibility of ONB-CBCT.

Results:As the beam becomes narrower, more and more scatter signals are eliminated. For the case with a bowtie filter and using 15 ONBs, the maximum and the average intensity error due to scatter are below 20 and 10 HU, respectively.

Conclusion: ONB yields a narrowed exposure field at each snapshot and hence an inherently negligible scatter effect. Meanwhile, the individualized detector units guarantee high frame rate detection and hence a same large volume coverage as that in conventional CBCT. In summary, ONB-CBCT is a promising design to achieve high-quality CBCT imaging.

Funding Support, Disclosures, and Conflict of Interest: This study is supported in part by NIH (1R01CA154747-01)

Contact Email: