Related Experiment Video
Updated: Oct 26, 2026
Positron Emission Tomography-based Dose Painting Radiation Therapy in a Glioblastoma Rat Model using the Small Animal Radiation Research Platform
Published on: March 24, 2022
Automated offline adaptive radiotherapy through CBCT imaging: A multidimensional feasibility study of hypersight and
Ben R Archibald-Heeren1, Aidan Leong2, Nigel Middlebrook3
1Icon Group, Clinical Care, Brisbane, Queensland, Australia.
Background:
Offline adaptive radiotherapy (OffART) offers potential dosimetric and clinical advantages by adjusting planned dosimetry in response to systematic patient changes throughout treatment, however implementation is associated with an increased labor burden. Technological innovations in automated contouring, image registration and on-board imaging present the opportunity for an automated workflow that achieves the benefit of OffART with a reduced impact on staff workload.
Purpose:
To assess if the image quality of a ring-based cone beam computed tomography (CBCT) imaging system is sufficient to enable automated contouring, deformable image registration (DIR), and dose calculation, thereby facilitating the implementation of an automated offline adaptive workflow.
Methods:
The treatment images for 40 clinical patients were retrospectively analyzed. The cohort spanned across multiple disease sites, including breast, thorax, abdomen, and male pelvis. Patients were selected from a clinical department utilizing the Varian Halcyon HyperSight (HypS) imaging system, and another department utilizing the Varian TrueBeam iterative cone-beam CT (TB-iCBCT) imaging system. Automated processes, without human intervention, were completed using the IGRT images and compared with planning simulation CT (pCT) images. Images were assessed for region of interest truncation, automated contouring agreement, DIR accuracy, and dose calculation accuracy using qualitative visual assessment, dice similarity coefficient (DSC) contour agreement and gamma analysis for dose calculation.
Results:
HypS imaging demonstrated improvements in reduced structure truncation with a reduction of 8% of structures with major or complete truncation. 13/21 organs demonstrated improved automated contouring consistency with pCT in HypS relative to TB-iCBCT. Overall DSC results for DIR propagated contours were higher in the HypS cohort, partially due to TB-iCBCT image truncation and partially due to improved registrations on the HypS images. HypS imaging demonstrated stronger agreement with pCT reported tissue density in heterogeneous patient anatomy than TB-iCBCT imaging. Dose calculation gamma results for HypS were on average higher than TB-iCBCT when comparing to the pCT calculated dose, however patient specific anatomical changes resulted in variability in results. Both imaging systems scored high gamma passing rates for calculations on pelvis cases.
Conclusion:
Improved image quality and field of view on the HypS imaging system results in improved system performance for several of the processes within the workflow for offline adaptation. Dose calculation and tissue density representation is of sufficient accuracy to warrant use of the imaging for offline adaptive radiotherapy. Automated contouring and contour propagation accuracy improved with HypS in some scenarios but remains the major barrier to a truly automated offline adaptive workflow not requiring human intervention.
