Room: AAPM ePoster Library
Purpose: present an improved stoichiometric method to represent high-density materials in the extended Hounsfield units (EHU, 30,710 HU) calibration curve used in proton treatment planning system. Most treatment planning systems rely on standard stoichiometric calibration performed in the conventional HU (CHU, 3071 HU) range for dose calculations
Methods: Gammex 467 phantom with standard tissue insert plugs was scanned using a Siemens Somatom with a 120 kV X-ray tube voltage. Various tissue plugs with known density ranging from cortical bone to lung were used. In subsequent CT scans, metal plugs (Mg, Al, Ti, Fe, Cu, Mo, W) with high purity and known density were used to replace the standard plugs. CT images were reconstructed in CHU and EHU modes and imported into Raystation TPS to construct CHU and EHU calibration curves. Dose distributions were calculated with CHU and EHU methods for phantom and patient studies. Two clinical cases (prosthetic hip and breast expander) involving metal implants were selected. Dose volume histograms were used to assess dosimetric differences for targets and normal tissues.
Results: was no difference in overall image quality or HU values of the CT dataset using EHU or CHU reconstruction. There was also a strong agreement between CHU and EHU CT calibration curves within the CHU range. Proton dose computations on phantom and retrospective patient studies produced similar dosimetric results using CHU and EHU calibration curves. However, EHU-based computations produced improved dosimetric accuracy to CHU-based computations when proton beam traversed close to or through the high-density artifacts.
Conclusion: Proton dose computations using EHU and CHU were shown to be in good agreement on phantom and patient studies. EHU-based calculations produced more accurate and consistent proton plans over a range of metal types and geometries than plans performed with object or planner-contoured density overrides.
Not Applicable / None Entered.
Not Applicable / None Entered.