Use of Dual Source Dual Energy CT for Photon-Based Dose Calculation In Radiation Oncology – a Comparative Study Based on Virtual Monoenergetic and Conventional 120 Kvp Images
Abstract
Purpose
Virtual monoenergetic images (VMIs) generated from dual-energy CT (DECT) offer diagnostic advantages such as enhanced contrast and reduced metal artifacts; however, their role in radiation oncology treatment planning is less established. This study evaluates the impact of VMIs on dose calculations in the Eclipse (Varian) treatment planning system, using the conventional 120-kVp CT protocol as the clinical reference. It investigative whether VMIs produce differential effects on dose calculation accuracy when using the Anisotropic Analytical Algorithm (AAA) compared with Acuros XB (AXB).
Methods
A CIRS Electron Density phantom was scanned on a dual-source DECT simulator to obtain Hounsfield Unit (HU) curves for VMIs ranging from 40-190 keV. A Catphan 500 phantom was also scanned, with VMIs reconstructed at 40, 60, 80, and 190 keV. Using these HU curves and corresponding Catphan datasets, treatment plans were created in Eclipse for 6X and 15X photon beams, each delivering 100 MU. LDPE and Teflon inserts served as target volumes representing low- and high-density tissues, respectively. Dose-Volume histograms (DVHs) were evaluated using six metrics: minimum-dose, D90, mean-dose, D50, D10, and maximum-dose. Percent differences were calculated relative to 120 kVp plans and between AAA and AXB for corresponding image energies.
Results
VMIs at 60 and 80 keV produced DVHs most consistent with 120 kVp reference, with average percent differences below 1% for both materials and algorithms. Larger deviations occurred at 40 and 190 keV, particularly for 6X AXB plans, which exceeded 2%. AXB showed greater sensitivity to HU variations in high-density Teflon than AAA, reflecting algorithm-specific material modeling, while LDPE exhibited consistent algorithm-dependent behavior across beam energies.
Conclusion
DECT-derived VMIs show promise for radiation therapy planning, contingent on appropriate HU validation and further patient-based evaluation. VMI energy selection may be tailored to optimize dose accuracy or image quality, analogous to diagnostic imaging.