Encrypted login | home

Program Information

The Use of CBCT for Dose Calculations in H/N Patients: Impact of Limited FOV

B Rosen

B Rosen*, C Lee , K Brock , University of Michigan Health System, Ann Arbor, MI


SU-F-J-15 (Sunday, July 31, 2016) 3:00 PM - 6:00 PM Room: Exhibit Hall

Purpose: To investigate the impact of limited superior and inferior (SI) field-of-view (FOV) associated with on-board kV CBCT images for dose calculation.

Methods: First, a (10x10) cm² continuous arc 6 MV treatment plan was calculated for fourteen virtual water phantoms of varying SI thicknesses (5 to 15 cm) to replicate the FOV effect. In addition, 70 Gy VMAT plans from two H/N patients were calculated on their corresponding planning CT (pCT) and first fraction CBCT image volumes. For the phantom, the mean dose difference (full vs. limited FOV phantom) was calculated for each axial plane for each size phantom to determine the error with distance from the SI boundary, as a function of dose at the FOV edge. For the clinical cases, the absolute difference between CBCT and pCT dose was calculated for each voxel within the CBCT FOV and evaluated relative to CBCT SI boundary.

Results: Phantom calculations showed that mean error was less than 5% for all planes whenever the CBCT border received less than 35 Gy or, otherwise, in planes at least 1.5 cm inside the CBCT SI FOV. Maximum pCT-calculated doses differences in the axial plane of the superior CBCT FOV border was 4 Gy for Patient 1 and 55 Gy for Patient 2, and the corresponding maximum error in the same planes were 2% and 24%, respectively. Maximum pCT-calculated dose 1 cm inferior to the superior FOV border was 7 Gy and 72 Gy, with a corresponding error of 0.5% and 2.5%.

Conclusion: CBCT images may be used as a surrogate for full CT images with the following limitations: low dose (<50% of prescription dose) at the FOV edge, or greater than 1.5 cm material thickness between the CBCT border and the axial plane of interest.

Funding Support, Disclosures, and Conflict of Interest: This work was partially funded by NIH P01CA059827.

Contact Email: