Program Information
A Model of Baseline Shift to Improve Robustness of Proton Therapy Treatments of Moving Tumors
K Souris1*, A Barragan Montero1 , E Sterpin1,2 , D Di Perri1 , X Geets1 , J Lee1 , (1) Universite catholique de Louvain, Brussels, Belgium, (2) KU Leuven, Leuven, Belgium
Presentations
MO-FG-CAMPUS-TeP3-1 (Monday, August 1, 2016) 5:30 PM - 6:00 PM Room: ePoster Theater
Purpose: The shift in mean position of a moving tumor, also known as “baseline shift”, has been modeled, in order to automatically generate uncertainty scenarios for the assessment and robust optimization of proton therapy treatments in lung cancer.
Methods: An average CT scan and a Mid-Position CT scan (MidPCT) of the patient at the planning time are first generated from a 4D-CT data. The mean position of the tumor along the breathing cycle is represented by the GTV contour in the MidPCT. Several studies reported both systematic and random variations of the mean tumor position from fraction to fraction. Our model can simulate this baseline shift by generating a local deformation field that moves the tumor on all phases of the 4D-CT, without creating any non-physical artifact. The deformation field is comprised of normal and tangential components with respect to the lung wall, in order to allow the tumor to slip within the lung instead of deforming the lung surface. The deformation field is eventually smoothed in order to enforce its continuity. Two 4D-CT series acquired at 1 week of interval were used to validate the model.
Results: Based on the first 4D-CT set, the model was able to generate a third 4D-CT that reproduced the 5.8 mm baseline-shift measured in the second 4D-CT. Water equivalent thickness (WET) of the voxels have been computed for the 3 average CTs. The root mean square deviation of the WET in the GTV is 0.34 mm between week 1 and week 2, and 0.08 mm between the simulated data and week 2.
Conclusion: Our model can be used to automatically generate uncertainty scenarios for robustness analysis of a proton therapy plan. The generated scenarios can also feed a TPS equipped with a robust optimizer.
Funding Support, Disclosures, and Conflict of Interest: Kevin Souris, Ana Barragan, and Dario Di Perri are financially supported by Televie Grants from F.R.S.-FNRS.
Contact Email: