Degree

Doctor of Philosophy (PhD)

Department

Department of Physics and Astronomy

Document Type

Dissertation

Abstract

Purpose: Local tomography reconstruction is achievable with EPID images acquired during VMAT and was named as VMAT-CT. However, it did not gain popularity due to multiple limitations and technical challenges. The goal of this study was to extend VMAT-CT concept, generate complete 3D or 4D CT images and dose, track and adapt VMAT plan based on updated images and dose.

Methods: We considered collimator angle and removed blurred areas in EPID images for VMAT-CT reconstruction to reduce artifacts and improve image quality. VMAT-CT+ images were generated by fusing VMAT-CT and planning CT using rigid or deformable registration. For 4D SBRT VMAT, breathing signal was extracted from EPID images which were sorted into four phases, and dose was calculated in each phase and registered to the mean position planning CT to generate 4D composite dose. Doses based on VMAT-CT+ and CBCT were compared for phantoms and real patients. When prescription dose was not met for PTV, re-planning was demonstrated on the phantoms. Possible uncertainties were also evaluated.

Results: Tracking based on VMAT-CT+ was accurate and superior to that based on CBCT since VMAT-CT+ can detect changes after setup. VMAT-CT could accurately detect phantom deformation and/or change of breathing pattern, and re-planning based on VMAT-CT could restore target coverage in both 3D and 4D cases. For the real patients, dose based on VMAT-CT agreed well with that based on CBCT acquired on the same day. The impact of uncertainties on dose was minimal for both 3D and 4D cases.

Conclusion: 3D and 4D tracking and adaptation of VMAT based on VMAT-CT are feasible. Our study can be incorporated into patients’ daily routine and has the great potential to increase the confidence of beam delivery, catch and remedy errors during VMAT.

Committee Chair

Zhang, Rui

Available for download on Thursday, March 16, 2023

Included in

Other Physics Commons

Share

COinS