Room: AAPM ePoster Library
Purpose: To investigate delivered dose estimation methods using weekly KV-CBCT to facilitate dose assessment for head-and-neck adaptive radiotherapy.
Methods: Weekly KV-CBCT images were acquired using the on-board imaging system for the setup of patients treated with Varian TrueBeam (Palo Alto, CA.) Six patients were retrospectively identified to ensure that the fields of view of all CBCTs contain the complete parotids. Online CBCT and planning CT registrations used for patient setup as well as any post-CBCT couch correction were retrieved from ARIA to align CBCT to the treatment position. Once the CBCT was aligned, delivered dose for the given CBCT was estimated by two
methods: M1) We deformed the planning CT to CBCT using the deformable image registration workflow in MIM (Beachwood, OH) and recalculated the dose with the deformed planning CT using the original plan; and M2) Planing dose is copied to the aligned CBCT. We compared the mean dose of the parotid and submandibular glands from the two methods in the 24 CBCTs (4 weeks per patient.)
Results: Mean dose of parotids estimated by M1 is 1.1(±1.3)% smaller than the mean dose estimated by M2 with 95% CI of [-1.5%, 0.7%] and mean dose of submandibular glands estimated by M1 is 0.4(±1.1) % smaller than the mean dose estimated by M2 with 95% CI of [-0.7%, -0.1%] The difference between M1 and M2 is not correlated to the volume shrinkage of parotids (Pearson’s r = 0.34, p=0.98.)
Conclusion: The difference of estimated delivered dose for salivary glands between with- and without- recalculation is within 2% and the difference is deformation-invariant. With proper patient setup alignment between CBCT and planning CT, the planning dose distribution itself is sufficient for speedy dose accumulation in head-and-neck adaptive radiotherapy.
Funding Support, Disclosures, and Conflict of Interest: This study is supported by a research agreement with Varian Medical Systems