Session Item

Daily CBCT for dose accumulation in head and neck (HN) radiotherapy (RT) is challenged by the limited field of view (FOV) of the CBCT, which excludes critical normal tissues and CTVs.  The goal of this work is to validate the use of CBCT padding to enable daily dose accumulation.

Planning CT (pCT), weekly CTs (wCT), and daily CBCTS were obtained for patients enrolled on an institutional review board approved clinical trial for adaptive RT. CBCT padding was achieved using ANACONDA DIR between the pCT and wCT and the CBCTs images. Only intensity information was used to drive the deformation. Outside of the CBCT FOV, where information to drive the DIR is not present, a smooth transition between the DVF in the region of the CBCT and a purely rigid registration at the boundary of the CT was achieved through iteratively increasing the regularization until a non-inverted DVF was achieved. Using the DVF, a padded CBCT was generated by augmenting the CBCT data with the transformed CT data. The DVF was used also for propagating the pCT contours onto the padded CBCT. To be able to compute dose on the images, the CBCT was corrected for shading artefacts and converted to HU using an algorithm available in the research version of RayStation. Dose was then calculated on the CBCT, pCt padded CBCT, and wCT padded CBCT. Qualitative evaluation was performed to assess the transition zone of the padded CBCT.  CT to CBCT DIR accuracy was assessed by evaluating the Dice Similarity Coefficient (DSC) of all structures. Clinical dose metrics were computed and compared between the same day wCT and the CBCT, CBCT padded with wCT, and CBCT padded with pCT.

CBCT padding has been evaluated for 42 images on 7 patients to date.  Qualitative evaluation demonstrated smooth transitions with minimal artifacts using the pCT and wCT for padding. DIR-based contour propagation resulted in DSC greater than intra-observer contour variation (0.8) for all structures (OARs, GTV, CTV, and PTVs) except cord (0.79), esophagus (0.79), brachial plexes (0.62), and cochleas (0.60). The average difference between the same day wCT and the CBCT padded with wCT was -0.11 cGy (0% of planned fraction dose, SD 0.99 cGY); CBCT padded with pCT was -0.08 cGy (0%, SD 0.73 cGy); CBCT without padding was 3.75 cGy (0.25%, SD 2.03 cGy). Although padding with the wCT had a smaller dosimetric difference, time trends were evaluated using the Mann-Kendall test at a significance level of 0.05. None of the tests produced significant p-values, which points to the absence of trends in the dose differences over time between padding with the pCT or wCT.

DIR can be accurately performed between CT and CBCT. CBCT padding with planning or weekly CT images enables accurate dose calculations over the full FOV. Combined together, these tools enable full FOV dose accumulation over the course of treatment using CBCT to support adaptive RT in HN.