Click here to


Are you sure ?

Yes, do it No, cancel

Evaluation of the Deformable Registration of Two Commercial Software Regarding the Estimation of Inter-Fractional Dose Delivery Accuracy of Prostate VMAT Plans

N Papanikolaou1*, J Hammers2 , S Stathakis3 , D Baltas4 , R Chen5 , S Das6 , P Mavroidis7 , (1) Ut Health San Antonio, San Antonio, TX, (2) ,Chapel Hill, NC, (3) University Of Texas Health, San Antonio, TX, (4) university Freiburg, Freiburg, ,(5) UNC Health Care, Chapel Hill, NC, (6) University of North Carolina, Chapel Hill, NC, (7) university North Carolina, Chapel Hill, NC


(Sunday, 7/29/2018) 3:00 PM - 6:00 PM

Room: Exhibit Hall

Purpose: To estimate the inter-fractional dose variation in treatment delivery of prostate cancer patients treated with VMAT. To compare the performance of MIM and Velocity in determining the fractional delivered doses to bladder and rectum.

Methods: Ten prostate cancer patients were treated with VMAT and daily CBCT scans were acquired. The planning dose was rigidly transferred to the CBCTs. The structures of bladder and rectum were either contoured manual or transferred from the planning CT using two different registration methods (contour-based or intensity-based deformable registrations (DIR)). For one of the patients, weekly CBCTs were selected and the fractional dose distributions to bladder and rectum were calculated per software and DIR method. The results were compared dosimetrically and radiobiologically. For the calculation of the NTCP values, the Relative Seriality model was used. For bladder the model parameters were D50=69.56Gy, γ=1.7, s=0.35, whereas for rectum they were D50=69.75Gy, γ=2.3, s=0.84.

Results: Comparing the manual contour DVHs against the plan DVHs we observe a standard deviation for NTCP and mean dose of 5.0% and 8.4Gy for bladder and 3.1% and 4.6Gy for rectum. The respective values from the comparison between the manual contour DVHs and the DVHs from the contour-based DIR of Velocity were 1.9% and 4.1Gy for bladder and 1.5% and 1.5Gy for rectum. The same comparison for the contour-based DIR of MIM gave 0.9% and 1.0Gy for bladder and 0.9% and 1.2Gy for rectum. Repeating this comparison for the intensity-based DIR of MIM showed deviations of 5.6% and 9.0Gy for bladder and 2.7% and 3.4Gy for rectum.

Conclusion: The dosimetric and radiobiological analysis showed considerable interfractional variations compared to the treatment plan for bladder and to a lesser extent for rectum. The contour-based DIR performed well and similarly in MIM and Velocity. The intensity-based DIR performed worse than the contour-based DIR.


Image-guided Therapy, Cone-beam CT, Shape Deformation


IM/TH- Image registration : CT

Contact Email