Quantitative dual-energy CT imaging for cancer imaging and radiation therapy applications Optimal planning and delivery of proton therapy and brachytherapy dose distributions for cancer treatment require accurate knowledge of the radiological properties (photon cross sections and charged-particle stopping powers) of tissue near the target volume and organs at risk. For proton therapy, uncertainties in stopping power maps measured by conventional single-energy computed tomography (CT) imaging necessitate use of range uncertainty margins of 3.5% that hinder sparing of adjacent organs at risk, thereby compromising treatment effectiveness. Current 125I, 103Pd, and electronic brachytherapy planning practice ignore deviations of tissue composition from liquid water, giving rise to 10%-100% dose delivery errors. While model-based dose- calculation engines are available, methods for accurate in vivo characterization of tissue inhomogeneities are urgently needed to eliminate these large dose specification uncertainties. In principle, dual-energy CT can provide more accurate material composition information, but to date clinical implementation has been limited by sensitivity to CT noise and artifacts. We have developed iterative dual-energy x-ray CT image reconstruction algorithms, incorporating realistic models of CT acquisition physics and measurement noise. Our promising preliminary results demonstrate tissue cross section maps that are more accurate and have lower uncertainty than either post-reconstruction dual- energy CT mapping processes or state-of-the-art single-energy stoichiometric tissue property mapping. We hypothesize that treatment plans based on these more accurate in vivo tissue cross section will decrease dose delivery uncertainty and improve plan optimality, leading to measureable reductions in normal tissue complications relative to tumor control achieved. We propose to conduct a virtual clinical trial based on modified treatment plans to demonstrate these improvements. Our proposal consists of three specific aims: 1) To implement and validate clinically usable dual-energy CT cross-section mapping processes for conventional multi-slice CT imaging systems used for radiotherapy simulation and treatment verification; 2) To evaluate intra-organ and inter-patient variations in radiological quantities and utilize these data to evaluate dose-delivery errors for proton therapy and low-energy brachytherapy; 3) To conduct a prospective virtual clinical trial to assess the impact of dual-energy CT tissue property mapping on plan quality, dose-delivery accuracy, and simulated patient outcomes for locally advanced lung and head-and-neck cancer patients treated with proton therapy. Our team of experienced collaborators will quantify the achievable accuracy for clinical tissue composition obtained by the combination of dual-energy CT and advanced reconstruction algorithms, and estimate how incorporation of this information improves treatment planning and patient outcomes. This work will translate the benefits of new technologies closer to clinical deployment.

Public Health Relevance

We will combine advanced reconstruction software with data collected by dual-energy CT scanners to improve estimates of tissue composition accuracy for cancer tumors and surrounding organs. This information will be used to create more accurate radiation dose delivery in proton therapy and low-energy brachytherapy treatments, which will improve outcomes and reduce side effects for cancer patients.

Agency
National Institute of Health (NIH)
Institute
National Cancer Institute (NCI)
Type
Research Project (R01)
Project #
5R01CA212638-04
Application #
9939481
Study Section
Radiation Therapeutics and Biology Study Section (RTB)
Program Officer
Obcemea, Ceferino H
Project Start
2017-03-01
Project End
2022-02-28
Budget Start
2020-03-01
Budget End
2021-02-28
Support Year
4
Fiscal Year
2020
Total Cost
Indirect Cost
Name
Washington University
Department
Engineering (All Types)
Type
Biomed Engr/Col Engr/Engr Sta
DUNS #
068552207
City
Saint Louis
State
MO
Country
United States
Zip Code
63130
Han, Dong; Porras-Chaverri, Mariela A; O'Sullivan, Joseph A et al. (2017) Technical Note: On the accuracy of parametric two-parameter photon cross-section models in dual-energy CT applications. Med Phys 44:2438-2446