Program Information
Head-And-Neck Digital Phantoms for Geometric and Dosimetric Uncertainty Evaluation of CT-CBCT Deformable Image Registration
Z Shen1*, K Bzdusek2 , S Koyfman1 , P Xia1 , (1) The Cleveland Clinic, Cleveland, OH, (2) Philips, Fitchburg, WI
Presentations
SU-E-J-119 (Sunday, July 12, 2015) 3:00 PM - 6:00 PM Room: Exhibit Hall
Purpose: To evaluate geometric and dosimetric uncertainties of CT-CBCT deformable image registration (DIR) algorithms using digital phantoms generated from real patients.
Methods: We selected ten H&N cancer patients with adaptive IMRT. For each patient, a planning CT (CT1), a replanning CT (CT2), and a pretreatment CBCT (CBCT1) were used as the basis for digital phantom creation. Manually adjusted meshes were created for selected ROIs (e.g. PTVs, brainstem, spinal cord, mandible, and parotids) on CT1 and CT2. The mesh vertices were input into a thin-plate spline algorithm to generate a reference displacement vector field (DVF). The reference DVF was applied to CBCT1 to create a simulated mid-treatment CBCT (CBCT2). The CT-CBCT digital phantom consisted of CT1 and CBCT2, which were linked by the reference DVF. Three DIR algorithms (Demons, B-Spline, and intensity-based) were applied to these ten digital phantoms. The images, ROIs, and volumetric doses were mapped from CT1 to CBCT2 using the DVFs computed by these three DIRs and compared to those mapped using the reference DVF.
Results: The average Dice coefficients for selected ROIs were from 0.83 to 0.94 for Demons, from 0.82 to 0.95 for B-Spline, and from 0.67 to 0.89 for intensity-based DIR. The average Hausdorff distances for selected ROIs were from 2.4 to 6.2 mm for Demons, from 1.8 to 5.9 mm for B-Spline, and from 2.8 to 11.2 mm for intensity-based DIR. The average absolute dose errors for selected ROIs were from 0.7 to 2.1 Gy for Demons, from 0.7 to 2.9 Gy for B-Spline, and from 1.3 to 4.5 Gy for intensity-based DIR.
Conclusion: Using clinically realistic CT-CBCT digital phantoms, Demons and B-Spline were shown to have similar geometric and dosimetric uncertainties while intensity-based DIR had the worst uncertainties. CT-CBCT DIR has the potential to provide accurate CBCT-based dose verification for H&N adaptive radiotherapy.
Funding Support, Disclosures, and Conflict of Interest: Z Shen: None; K Bzdusek: an employee of Philips Healthcare; S Koyfman: None; P Xia: received research grants from Philips Healthcare and Siemens Healthcare
Contact Email: