Room: AAPM ePoster Library
Purpose: To investigate the feasibility of using average intensity projection (AIP) of four-dimensional (4D) in-treatment cone-beam computed tomography (CBCT) for patient setup verification.
Methods: A phantom study was performed to evaluate the reproducibility of a lung target position visualized in AIP of in-treatment CBCT images. An XSight lung tracking phantom was used, in which a spherical target (2.5 cm diameter) moves in the superior-inferior direction with 3 cm excursion and 5 seconds cycle. The phantom was scanned using an Elekta XVI imager and acquired projections were reconstructed into 10-phase 3D images with a voxel size of 2 mm in all directions. CBCTs were acquired without and with delivery of SBRT plans: pretreatment vs. in-treatment CBCTs. For acquisition of in-treatment CBCTs, 16 volumetric-modulated arc therapy (VMAT) plans using unflattened x-ray beams were delivered; the monitor units (MUs) of the beam concurrently delivered were 2062.6 ± 1221.2 MU (153.8-4524.9). The lung target in each of the AIPs of the in-treatment CBCT images was rigidly aligned to that of the pretreatment CBCT using RayStation (translation only). The Pearson correlation analysis was performed between the MUs and target registration offset using MATLAB.
Results: The absolute mean and standard deviation of the resulting translation in the superior-inferior direction were 0.4 ± 0.5 mm (-1.4-0.5), indicating patient setup verification can be performed with a sub-pixel accuracy using the AIP of the in-treatment CBCTs. The target registration accuracy was not affected by the MUs delivered by the VMAT plans as no statistically significant correlation was found between the MUs and the registration offset. The Pearson correlation coefficient and corresponding p-value were 0.18 and 0.51.
Conclusion: This study demonstrates that a moving target can be reproducibly scanned by in-treatment CBCT, so that the resulting AIP can be successfully used for verification of lung target during SBRT.
Funding Support, Disclosures, and Conflict of Interest: This work was supported by a research agreement with Elekta Limited.