Author: Wen C. Hsi, Pouya Sabouri, Zhong Su 👨🔬
Affiliation: University of Arkansas for Medical Sciences, Department of Radiation Oncology, University of Arkansas for Medical Sciences (UAMS) 🌍
Purpose:
Traditional single-energy CT (SECT) contrast scans cannot be used for proton dose calculations due to significantly higher HU caused by iodine. Dual-layer dual-energy CT (DL-DECT) can directly obtain the electron density (ED), and effective atomic number (Zeff) of the imaged object to derive the proton stopping-power ratio (PSPR) dataset for proton dose calculations. This study is to evaluate the dosimetric impact of contrast on direct PSPR dose calculations and feasibility of using only DL-DECT contrast scan for treatment planning.
Methods:
Total 16 patients have both true non-contrast (TNC) and contrast scans from a DL-DECT. The clinical proton plans were performed on the TNC plans with HU-PSPR calibration curves. PSPR datasets from the TNC and contrast scans were obtained from the DECT datasets of ED and Zeff with Bourque et al. method. The clinical plans were copied onto the PSPR datasets and recalculated. For dosimetric comparisons, the clinical goals were examined on dose calculations of PSPR datasets from both TNC and contrast scan and the results are presented.
Results:
For target covered by prescribed dose, plans differences are within 1.3% except one patient with difference of 3.4% between the TNC and Contrast PSPR datasets. For HN patients, the target coverage variation is -0.3%+/-0.9%. For important OARs, i.e., brainstem, cord, chiasm, optical nerves, the variations between TNC and Contrast plans are 5.3+/- 17.7cGy; for all OARs, the variations are 0.7+/- 32cGy. The maximum variations for cases of brain, lymphoma, and pelvis cases are 1.59% for CTV coverage and 20cGy for OARs.
Conclusion:
The direct PSPR conversion from the DL-DECT scans avoid the uncertainties and degeneracy of HU to PSPR conversion. The PSPR from contrast scan has minimal impact from iodine. Thus, single DL-DECT contrast scan generated PSPR dataset can be directly used for proton dose calculations with clinical acceptable uncertainties.