Program Information
Simultaneous Multi-Energy Cone Beam CT Reconstruction and Material Decomposition Realized On a Conventional Cone Beam CT Platform
C Shen1*, B Li1 , Y Lou2 , X Jia1 , (1) University of Texas Southwestern Medical Center, Dallas, TX, (2) The University of Texas Southwestern Medical Center, Dallas, TX
Presentations
WE-F-605-9 (Wednesday, August 2, 2017) 1:45 PM - 3:45 PM Room: 605
Purpose: Cone beam CT (CBCT) is currently the most widely used image-guidance tool in radiotherapy. However, its use is still mainly limited to patient positioning. This study aims at exploring the potential of the clinical CBCT platform in terms of performing multi-energy (ME) scans and reconstructing material-decomposed images to support advanced applications in radiotherapy, such as stopping power calculation for online adaptive particle therapy and identification of low-concentration contrast agent for image guidance.
Methods: ME-CBCT was realized on a Varian TrueBeam CBCT platform using a fast kVp-switching scanning scheme via its developer mode. The simultaneous CBCT reconstruction and material decomposition problem was formulated as an optimization problem to directly solve for x-ray attenuation coefficient, electron density relative to water (rED) and elemental composition (EC) of each voxel. The objective function used a least square term together with a tight-frame regularization to ensure fidelity between attenuation coefficient and projection measurement. Self-consistency between rED, EC and attenuation coefficient was imposed. The EC was further subject to a constraint that it is a sparse combination of the material’ ECs in a dictionary containing materials commonly seen in human body. The optimization problem was solved by a novel alternating-direction minimization scheme. We tested the proposed framework by experiments using a Gammex phantom.
Results: We realized the ME-CBCT scan with 80, 100 and 120 kVps. The mean relative error of rED and EC were 1.17% and 2.07%. The reconstructed ME-CBCT images of x-ray attenuation achieved ~4 times higher contrast-to-noise-ratio than those of FDK algorithm.
Conclusion: We have developed a novel and practical framework to realize ME-CBCT on a standard clinical CBCT system through a kVp-switching strategy. A new reconstruction method that simultaneously reconstruct ME-CBCT images and determine material compositions were developed and tested. The proposed framework may be applied to advanced applications in radiotherapy.
Contact Email: