Author: Chloe DiTusa, Panayiotis Mavroidis, Christopher W. Schneider, Sotirios Stathakis 👨🔬
Affiliation: Louisiana State University, Mary Bird Perkins Cancer Center, University of North Carolina 🌍
Purpose: To evaluate and compare dose calculation differences between Monaco and AdaptBox by TheraPanacea on AI-generated pseudo-CTs (pCTs) from a CBCT.
Methods: Dose calculations in water phantom were used to test the dose calculation accuracy of the AdaptBox. A head-and-neck phantom and four previously treated head-and-neck patients were selected for the study. CT scans and corresponding CBCTs were acquired, and pCTs with AI-generated contours were created. Two control plans were created for the phantom: a four-beam plan (box) and a single-beam plan using a 6MV photon beam. All patient plans were created using the same 6MV beam. Dose using the pCT were calculated using the collapsed cone convolution superposition in the TheraPanacea and recalculated in Monaco. Dose calculations in Monaco were performed with 1% statistical uncertainty per plan. Dose differences were evaluated using 3D gamma index analysis. DVH objectives were also compared to assess the delivered daily dose.
Results: Profiles and depth dose curves were within 1% agreement between the two systems for the water phantom. For the phantom and patient plans, the 3D gamma analysis demonstrated strong agreement between the dose distributions of both systems, with all patients achieving pass rates above clinically acceptable thresholds. Daily DVH analysis of the patient plans revealed 9 of 177 objectives in disagreement, with percent differences in pass/fail outcomes ranging from 0.3% to 27%. More specifically, higher dose differences (1.9%) for targets were reported by Monaco compared to TheraPanacea.
Conclusion: Overall, the AdaptBox system by Therapanacea shows good dosimetric agreement compared to the Elekta Monte Carlo even in the demanding cases of head and neck patients with many tissue heterogeneities. The dosimetric differences observed are expected due to the different dose calculation algorithms used by the two systems. These findings support the potential for TheraPanacea to be used reliably in clinical dose calculations.