Room: Exhibit Hall | Forum 6
Purpose: The accuracy of Hounsfield units (HUs) to electron densities (EDs) calibration curve is affected by CT scanning parameters. This study aims to demonstrate the feasibility of direct dose calculation using electron density images in commercial treatment planning system., ED image quality is evaluated and preliminary evaluation of the dosimetric impact is investigated.
Methods: ED and HU images were both generated from the same acquisition using a dual-layer CT system (Philips IQon Spectral CT) and imported separately into Varian Eclipse treatment planning system (V13.7.15). Intensity-modulated treatment plans of three cancer sites (brain, lung, head and neck) were designed based on targets and organs-at-risk delineated on HU images. All contours and treatment plans were then copied to the ED images. Doses were recalculated on ED images without modifying copied photon beam parameters.
Results: ED images can be successfully imported into commercial treatment planning systems. Beam hardening artifacts seen on iteratively reconstructed HU images were greatly reduced on ED images. The maximum difference found in the brain site was in the cochlea (0.27 Gy difference in mean dose). For the lung site, the maximum difference was in the PTV (0.22 Gy difference in max dose). For the head and neck site, the maximum difference was in the right lens (0.1 Gy difference in max dose).
Conclusion: It is feasible to directly calculate ED images using dual-layer CT and import into a treatment planning system for dose calculation. Preliminary results indicate dose distributions agreed well between ED and HU plans. Image artifacts were reduced and HU dependency on scan parameters was no longer a concern with ED images. A comprehensive study is ongoing to determine the site-specific dosimetric accuracy of direct ED planning and contrast enhancement effects on HU and ED.
Funding Support, Disclosures, and Conflict of Interest: This research was carried out under a research agreement with Philips Healthcare.
Not Applicable / None Entered.