Dosimetric Assessment of Treatment Using CBCT Images
A Dyess1*, E Poon2, J Seuntjens1, W Parker3, (1) McGill University, Montreal, QC, (2) Montreal General Hospital, Montreal, QC, (3) McGill Univ Health Center, Montreal, QCSU-E-J-208 Sunday 3:00:00 PM - 6:00:00 PM Room: Exhibit Hall
Purpose: To evaluate target coverage for five breast patients receiving boost treatment to the tumor bed by calculating the daily dose on cone beam computed tomography (CBCT) images and utilizing deformable image registration techniques.
Methods: The daily dose is calculated on pretreatment CBCT images using the same beam configuration as the original volumetric modulated arc therapy (VMAT) plan. Calculations were done with two different isocenter positions according to: (1) the initial patient setup and (2) the shifts applied for treatment based on CBCT verification. The daily doses are deformed and accumulated onto the planning CT using commercially available deformable image registration software. The dose distribution is compared to the original distribution and tumor and PTV coverage is evaluated for both situations (initial and shifted positions). The deformation accuracy is evaluated by calculating the change in centroid location and the Dice coefficient of deformed contours.
Results: The tumor bed is adequately covered regardless of the treatment position. The average dose received by 98 % (D98) of the tumor bed volume differs from the original plan by +1.6 % and -0.2 % for the shifted and initial positions respectively. However, when dose is accumulated in the initial setup position PTV coverage is lost; the average D98 for the PTV changes by -15.8 % and -26.9 % for the shifted and initial positions respectively. The average change in centroid location is 0.43 mm and 1.53 mm for the left and right lung contour respectively. The Dice coefficient for the left and right lung is 0.94 and 0.95 respectively.
Conclusions: The margins used to define the PTV are sufficient to ensure tumor bed coverage for the given positioning variability. We are also confident in the deformation used to deform and accumulate dose based on deformed contour comparison.