Monte Carlo (TOPAS) characterization of a proton therapy OBI: Automated source parameter optimization and analysis of CBCT acquisition dynamics
Medical Physics · 2026-11 · 53 (11)
DOI: 10.1002/mp.70753 ↗Kurz gefasst
Fachartikel in der Zeitschrift Medical Physics, online erschienen am 11.11.2026. Verfasst von Wei Chuen Tan und 9 weiteren Autor:innen. Beteiligte Einrichtung: Division of Radiation Oncology, National Cancer Centre Singapore, Singapore, Singapore. Hauptthemen laut MEDLINE: Monte Carlo Method, Cone-Beam Computed Tomography, Proton Therapy, Image Processing, Computer-Assisted. Die Arbeit nennt 37 Literaturangaben.
Abstract (Originalsprache)
Background
The cumulative radiation burden from daily cone‐beam CT (CBCT) imaging is a significant concern. While Monte Carlo (MC) simulation provides the gold standard for patient‐specific dosimetry, clinical implementation is hindered by the complexity of characterising proprietary X‐ray source parameters. Furthermore, the dosimetric properties of the Hitachi ProBeat on‐board imager (OBI), which utilizes a fixed frame rate (FFR) acquisition logic, remains uncharacterized in the literature.
Purpose
To develop and validate a robust indirect characterization methodology that automates MC beam tuning using standard QA measurements, and to utilize this model to quantify the dosimetric impact of the Hitachi OBI's acquisition dynamics.
Methods
A TOPAS (Tool for Particle Simulation) MC model of the OBI was constructed through indirect characterization. An automated derivative‐free optimization framework was developed to tune source parameters (spectrum, filtration, focal spot) by minimizing the loss against 90 static projection measurements acquired on a standard CT dose index (CTDI) phantom. A global scaling factor was used to scale simulated dose to physical dose. The model was validated against radiophotoluminescent dosimeter (RPLD) measurements in an ATOM 701 adult male anthropomorphic phantom for head and thorax protocols. Finally, the validated model was used to isolate the dosimetric overhead introduced by the system's fixed frame rate (FFR) logic compared to a theoretical fixed angular interval (FAI) acquisition.
Results
Optimization of beam parameters through indirect characterization successfully converged, yielding a set of simulation parameters and a general scaling factor function. In the anthropomorphic validation, the model predicted central organ doses with a mean absolute percentage error (MAPE) of 3.4%, with all in‐field validation points falling within 10% of physical measurements. Analysis of the gantry acquisition kinematics identified up to 9.4% additional projections that were not utilized during reconstruction, and up to 23.6% additional projections compared to a theoretical FAI acquisition. Simulations indicate dose hotspots approximately 10 mGy in bone structures due to kinematic oversampling at during acceleration/deceleration phases in FFR compared to FAI acquisition.
Conclusions
A high‐fidelity Monte Carlo beam model can be successfully derived solely from standard CTDI measurements via automated indirect characterization, achieving a 3.4% MAPE in anthropomorphic phantom validation. This accessible framework provides a practical alternative to specialized direct beam measurements, significantly lowering the barrier to clinical implementation. Applying this tool to the Hitachi OBI quantified specific dosimetric variations inherent to its kinematic acquisition logic, identifying theoretical avenues for integral dose reduction. This facilitates the transition to accurate, patient‐specific dosimetry, empowering clinics to evaluate risk‐adapted IGRT strategies and better track cumulative imaging doses.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Division of Radiation Oncology, National Cancer Centre Singapore, Singapore, Singapore.
- Department of Physics, National University Singapore, Singapore, Singapore.
- School of Physical and Mathematical Sciences, Nanyang Technological University, Singapore, Singapore.
- Department of Radiation Oncology, The University of Osaka Graduate School of Medicine, Osaka, Japan.
- Duke-NUS Medical School, Singapore, Singapore.
Zitieren
APA
Tan, W. C., Lee, K. H., Fu, K., Lew, K. S., Koh, W. Y. C., Chua, C. G. A., Yagi, M., Park, S. Y., Lee, J. C. L., & Tan, H. Q. (2026). Monte Carlo (TOPAS) characterization of a proton therapy OBI: Automated source parameter optimization and analysis of CBCT acquisition dynamics. Medical Physics, 53(11). https://doi.org/10.1002/mp.70753
Vancouver
Tan WC, Lee KH, Fu K, Lew KS, Koh WYC, Chua CGA, et al. Monte Carlo (TOPAS) characterization of a proton therapy OBI: Automated source parameter optimization and analysis of CBCT acquisition dynamics. Medical Physics. 2026;53(11). doi:10.1002/mp.70753
BibTeX
@article{Tan2026,
author = {Tan, Wei Chuen and Lee, Kang Hao and Fu, Kailin and Lew, Kah Seng and Koh, Wei Yang Calvin and Chua, Clifford Ghee Ann and Yagi, Masashi and Park, Sung Yong and Lee, James Cheow Lei and Tan, Hong Qi},
title = {Monte Carlo (TOPAS) characterization of a proton therapy OBI: Automated source parameter optimization and analysis of CBCT acquisition dynamics},
journal = {Medical Physics},
year = {2026},
volume = {53},
number = {11},
doi = {10.1002/mp.70753}
}
EUVIMED