Development of a dual-energy CBCT calibration method for accurate electron density estimation in preclinical
I-Chun Cho1, Tanaporn Buangklang2, Chien-Hau Chu3
1Research Center for Radiation Medicine, Chang Gung University, No.259, Wenhua 1st Rd., Guishan Dist., Taoyuan City, 333323, Taiwan; Radiation Research Core Laboratory, Linkou Chang Gung Memorial Hospital, No.15, Wenhua 1st Rd., Guishan Dist., Taoyuan City, 333011, Taiwan.
Abstract:
In preclinical animal studies for proton therapy, Computer Tomography (CT) imaging is essential for planning and proton dose calculation, which requires accurate conversion of Hounsfield Unit (HU) to stopping power ratio (SPR). While single-energy CT (SECT) is traditionally used, dual-energy CT (DECT) offers improved accuracy in estimating SPR. This study-the first animal dual-energy cone-beam CT (CBCT) study for proton therapy-developed a DECT calibration method for a small-animal CBCT system using a CIRS 062M electron density phantom scanned at 40 and 60 kVp. SECT calibration applied a stoichiometric method at 60 kVp, while DECT followed Saito's method (2017). Accuracy was assessed by comparing measured SPR values to theoretical predictions from the Bethe formula, using root mean square error (RMSE) as the metric. Results showed that both SECT and DECT enabled HU to SPR calibration, but DECT achieved better accuracy for soft and bone tissues, with RMSEs of 0.25% and 1.08%, respectively, compared to 0.87% and 1.44% for SECT. However, for high-density materials (relative electron density >2), SECT outperformed DECT, showing a lower RMSE (4.19% vs. 5.42%). These findings support the application of DECT in small-animal CBCT for improved electron density estimation and more accurate preclinical proton therapy dose planning.
