Click here to


Are you sure ?

Yes, do it No, cancel

Voxel-Specific Characterization of An Anthropomorphic Phantom as the Ultimate Ground Truth to Evaluate the Accuracy of Various Imaging Methods of Proton Therapy

C Chang*, J Harms, S Charyyev, J Zhou, X Yang, L Lin, Emory Proton Therapy Center, Atlanta, GA


(Sunday, 7/12/2020)   [Eastern Time (GMT-4)]

Room: AAPM ePoster Library

Purpose: from CT HU to proper material densities and compositions is the dominant component of dosimetric uncertainty in Monte Carlo-based proton treatment planning systems. Establishment of a voxel-specific characterization using an anthropomorphic phantom can be used as the ground truth to evaluate the magnitude of inaccuracy of CT imaging protocols due to beam hardening or material derivation.

Methods: material density and composition of CIRS M701 Adult Phantom were provided by the vendor. All 39 slabs of 25-mm thickness were irradiated with 208 MeV proton single spots (4-mm sigma) for large uniform regions, and a 2-mm diameter pinhole was used to collimate the beam for small uniform regions. Residual range was measured with a multi-layer ion chamber. Thicknesses of each material were obtained from CT images if multiple materials are involved to ensure the accuracy of relative stopping power (RSP) derivation from ground truth. Single-energy and dual-energy CT (SECT and DECT) images were acquired with a Siemens Twinbeam CT scanner.

Results: values for M701 phantom’s lung, soft tissue, brain, and bone are 0.201, 1.040, 1.051, and 1.474, and the relative differences of RSP between measurement and Bethe-Bloch calculation based on CIRS compositions are 1.3%, 0.1%, 0.2%, and 0.7%. RSP comparison between SECT and measurement shows ~4% underestimation for inner bones in pelvic region, and the RSP map derived from DECT parameters indicate residual uncertainties, likely from the compositional difference in CIRS bone from that assumed in syngo.via’s derivation of DECT parameters, indicating the need of protocol to be developed to derive the mass density and effective Z for artificial materials.

Conclusion: characterization of an anthropomorphic phantom is established. The method can be used to determine the tissue and location dependent uncertainties of SECT and identify the residual uncertainties from current DECT approaches.


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email