Room: AAPM ePoster Library
Purpose: Implementation, commissioning and validation of a dual energy CT (DECT) Varian Eclipse (application programming interface) API script for calculation of stopping power ratios (SPRs) and proton dose.
Methods: Using Siemens Definition Edge CT scanner, the GAMMEX phantom was scanned using sequential DECT at 140 kVp and 80 kVp. The calculated electron density (CT_rho) and effective atomic number (CT_Zeff) images was imported into Eclipse. The reference stopping power ratio (SPR) of the tissue equivalent plugs were measured using both proton irradiation as well as calculated using the Bethe-Bloch equation with the material composition data. The accuracy of SPRs calculated using the DECT API script was compared with the reference SPRs as well as single energy CT (SECT). Plan optimization was first performed using SECT and the DECT script was used to calculate DECT SPR and the forward calculated DECT dose distribution.
Results: The manually calculated values of SPRs matches the corresponding values from Eclipse DECT API across all effective Z values. Comparison of API and theoretical SPRs shows variance below 1.1 % across all tissue equivalents plugs except lung plugs, LN300 and LN450 (20 % difference). Comparison of SPR from SECT with theoretical calculations show variance of more than 4% for several plugs. Excluding the lung plugs, the average differences between measured values and the DECT is 0.9% and is 2% for SECT values.
Conclusion: The Eclipse DECT API is a streamlined tool for generating the SPRs values based on the DECT images allowing for direct comparison of SECT based dose calculation. The comparison of the SPRs generated by this tool and the theoretical calculations and measured values show good agreement for most tissue equivalent material. Through the use of the API script, clinical implementation of DECT can be performed in a convenient way that can lead to widespread adoption.
Funding Support, Disclosures, and Conflict of Interest: Received research funds from Varian
Not Applicable / None Entered.
Not Applicable / None Entered.