Program Information
Potential Implications of CT-To-Density Relationships On Dose Calculations
S Mossahebi*, M Guerrero , K Prado , K Langen , B Yi , S Chen , University of Maryland School of Medicine, Baltimore, MD
Presentations
SA-B-BRA|B-8 (Saturday, March 18, 2017) 10:30 AM - 12:30 PM Room: Ballroom A|B
Purpose: We investigate the effect of CT-to-density relationships and their associated heterogeneity corrections on dose calculation for large depths using Mobius3D and two treatment planning systems (TPS): RayStation (v.4.5) and Eclipse (v.11).
Methods: Measurements were made to quantify potential dose-calculation differences resulting from different CT-to-density tables. The TPS CT-to-density was generated from HU measurements made using commercial phantom (Gammex) on multiple CT scanners. The Mobius3D CT-to-density relationship was obtained from a bilinear fitting performed by the vendor. The dose calculations of TPSs and Mobius3D were compared with ion chamber measurements in multiple SAD setups at multiple depths using both water and plastic water phantoms. For water setup comparisons, identical mass density (1g/cm³) was used to calculate dose for both TPS and Mobius3D. For plastic-water setup comparisons, mass densities were obtained from the CT-to-density relationships. For each depth the TPSs and Mobius3D dose calculations were compared to the measurements.
Results: CT-to-density relations of TPS and Mobius3D showed up to 5% density difference for materials with -100≤HU≤100. In water, agreements within 1% and 1.5% were found between TPSs and Mobius3D dose calculations respectively with ion chamber measurements. Larger differences were observed in the plastic water phantom (HU≈60), however. The Mobius3D dose calculations agreed with ion chamber measurements within 2%, whereas differences of up to 6% were observed in the TPS data comparison. These results show that the magnitude of differences between TPS and Mobius3D calculations could arise from differences in heterogeneity corrections of calculated mass densities.
Conclusion: CT-to-density uncertainties could produce dose calculation differences exceeding 5% for large depths. Bilinear fitting of CT-to-density data seems to show better dose calculation agreement with measurements for low HU materials at large depths, than do those calculations that are based explicitly on CT-to-density relationships measured with commercial phantoms. This is currently under more rigorous investigation.
Contact Email: