Evaluating a CBCT Correction Algorithm for Adaptive Radiotherapy

IntroductionTo increase efficiency of adaptive radiotherapy (ART), we tested a cone beam computed tomography (CBCT) correction algorithm to evaluate the feasibility of utilizing daily CBCTs for treatment planning.MethodsA lung phantom was scanned with a CT and CBCT on two different linacs. The CBCTs were processed through a correction algorithm in the treatment planning system (TPS). The algorithm reduces artifacts and adjusts image intensity to more closely match the planning CT, to generate corrected CBCTs. Voxels outside the CBCT field of view (FOV) are replaced with voxels from the planning CT. A treatment plan was first generated on the CT, then recalculated on the corrected CBCTs. The same workflow was followed for seven previously adapted head and neck and seven sarcoma patients. Each patient's adaptive plan was recalculated on the corrected CBCTs. Dose differences were analyzed for these plans using a 3%/2 mm gamma analysis.ResultsBoth Ethos and TrueBeam CBCT plans on the phantom had high matching dose per voxel according to gamma analysis. After corrections of some registration errors, all 14 plans achieved gamma passing rate above 95% (3%/2 mm).ConclusionsThe CBCT correction algorithm demonstrates potential to reduce the need for re-simulation and enable faster offline adaptive planning without sacrificing dose calculation accuracy.