(Sunday, 7/29/2018) 1:00 PM - 1:55 PM
Room: Room 202
Purpose: To evaluate the potential of multi-energy CT to reduce proton beam range uncertainties in a full Monte Carlo environment.
Methods: A realistic and controlled ground truth patient geometry is created from a real patient pelvis scan, using contours made by an expert and tabulated tissue data. Virtual CT images of the ground truth patient are generated for single-, dual- and multi-energy CT (SECT, DECT and MECT). Images are generated in two contexts: A) theoretical CT numbers (best-case scenario) and B) reconstructed polyenergetic sinograms without artifact corrections (worse-case scenario). Post-reconstruction Poisson noise is considered in both cases. SECT images are converted in MC inputs using the method of Schneider et al. (2000), while the Bayesian eigentissue decomposition (BETD) of Lalonde et al. (2017) is used for the DECT and MECT images. Mono-energetic proton pencil beams are simulated from seven different angles around the patient using TOPAS-GEANT4, with an energy adjusted to reach the center of the prostate. Dose distributions and percentage depth dose curves are compared to the ground truth to compare the performance of the methods.
Results: In case A, using idealistic CT numbers only affected by noise, small benefits of DECT and MECT over SECT are apparent. Root mean squared (RMS) errors of 0.78 mm, 0.50 mm and 0.48 mm are observed for SECT, DECT and MECT respectively. For case B, MECT noticeably reduces the error on proton range compared to both SECT and MECT. RMS errors of 2.64 mm, 1.04 mm and 0.64 mm are obtained for SECT, DECT and MECT respectively.
Conclusion: The potential of MECT to reduce proton range uncertainties in the context of MC dose calculation is demonstrated. Simulated images in two limiting cases suggest that the benefits of MECT over DECT or SECT are important in the presence of imaging artifacts.
Keywords
Not Applicable / None Entered.
Taxonomy
Not Applicable / None Entered.
Contact Email
