Room: Exhibit Hall | Forum 5
Purpose: To investigate the feasibility of using synthetic CT generated from CBCT of the day during the course of proton therapy for monitoring changes in water equivalent thickness (WET) and for adaptive planning.
Methods: An anthropomorphic head phantom was scanned with CT and CBCT with and without a 10 mm water equivalent bolus applied around chin and neck areas to simulate body circumference changes from weight gain. CT images of the phantom without bolus, simulating the planning scan, were deformed using MIM software (normalized intensity-based deformable image registration) to CBCT images of the phantom with bolus, simulating the image guidance scan on a treatment day. The deformed CT scan (synthetic CT) was compared to CT images with the bolus, serving as the ground truth. The deformation accuracy was independently verified with ImSimQA software using a set of metrics recommended by the TG-132 report. Comparisons were made between the synthetic CT and the ground truth CT in terms of dose distributions and WET.
Results: Comparing to the planning CT, synthetic CT of the day detected changes in WET by 10.8Â±0.9 mm (mean Â± standard deviation) and a decrease in CTV D95 from 98.5% to 61% of the prescribed dose. The difference in WET between the ground truth CT and the synthetic CT was only 1.0Â±0.2 mm. The CTV D95 for the re-optimized plan based on ground truth CT and the recalculated dose on synthetic CT were both 98.4%. Similarity in WET and target dose suggested that the synthetic CT provides a good alternative to high quality re-scan CT for estimating the impact of body circumference changes.
Conclusion: Our experiments demonstrated that synthetic CT generated by deforming the planning CT to CBCT of the day is useful for evaluating and adapting to changes in WET and target coverage in proton therapy.