Relative Electron Density Phantom Comparison
B Rasmussen1*, K Chu1, S Tong2, (1) Marquette General Hospital, Marquette, MI, (2) Inova Alexandria Hospital, Alexandria, VASU-E-T-530 Sunday 3:00:00 PM - 6:00:00 PM Room: Exhibit Hall
Purpose: Modern treatment planning systems require lookup tables to convert Hounsfield Units (HU) to relative electron density (RED) for use in heterogeneity corrections during dose calculations. The purpose of this work is to illustrate the impact of using different model CT phantoms to determine HU to RED curves for treatment planning.
Methods: A GAMMEX model 467 tissue characterization phantom and a CATPHAN model 500 multi-purpose CT phantom were imaged using CT scanners in four different cancer centers and the HU to RED curves derived from each phantom were imported into an Eclipse 8.10 treatment planning system. Dose calculation were performed on a heterogeneity phantom and then compared to measurements. A comparison of isodose and DVH were performed by calculating 3D and IMRT plans onto identical CT datasets with different HU to RED curves to determine the clinical significance. Analysis based on effective atomic number of the phantom inserts was also performed.
Results: The HU to RED curves from the GAMMEX and CATPHAN phantoms were found to be reasonably self-consistent across the different CT scanners. However, observable differences for higher density materials were observed between the two phantom models. The differences for the larger HU values can be attributed to the effective atomic number of the materials. In kV range of a CT scanner x-ray spectrum, photon interactions are partially due to the photoelectric effect which has a larger dependence on atomic number than Compton scatter which depends most directly on electron density.
Conclusion: The HU to RED curve is more dependent on the phantom model than CT scanner. The HU to RED curve from the GAMMEX phantom produced better agreement between Eclipse AAA calculations and measured dose distributions on a heterogeneity phantom than that from the CATPHAN. However, DVH and isodose data on patient plans show small differences for common treatment sites.