4D cone-beam CT image reconstruction of Varian TrueBeam v1.6 projection images for clinical quality assurance of stereotactic ablative radiotherapy to the lung

• Main Author: Beaudry, Joel
• Language: English
• Published: University of British Columbia 2015
• Online Access: http://hdl.handle.net/2429/52814

On-board cone-beam computed tomography (CBCT) imaging integrated with medical linear accelerators offers a viable tool for tumor localization just prior to radiation treatment delivery. However, the exact tumor location during treatment is not well-defined due to respiratory motion. This is taken into account during treatment planning by adding margins to the visible tumor volume defining the high dose region. The respiratory motion used to optimize the treatment plan is not guaranteed to be reproducible on the day of treatment, suggesting that the high dose region may not fully contain the tumor at all points of its trajectory during treatment. In this thesis, to image the tumor at the different portions of the breathing cycle, CBCT projections were binned by the respiratory signal at their time of acquisition. Reconstructing each bin created a 3D image depicting the tumor at one point of its trajectory. Combining the binned reconstructions added in a temporal component, defining a 4D-CBCT. 4D-CBCT reconstructions were performed on 6 stereotactic ablative radiotherapy (SABR) lung cancer patients. Imaging was performed using the Varian TrueBeam (v1.6) and respiratory information was captured with the infra-red camera-based Varian real-time position management (RPM) system. Both analytical and iterative reconstruction algorithms, and image quality metrics were used for a comparative study. Tumor motion was measured by tracking the visible tumor volume centroid from each 4D-CBCT image. The high dose regions defined during treatment planning were compared to the 4D-CBCT tumor volume during its trajectory using an overlap metric to determine if the tumor remained confined to the treatment volume, or not. 4D-CBCTs were found to be well reconstructed using iterative methods. When viewed sequentially the 4D-CBCT images visibly show tumor motion following a sinusoidal-like behavior. Examination of the tumor motion and overlap metric verify that the margins currently used to define the high dose region fully encompass the tumor during all times of its trajectory, i.e 100% overlap within error. The results indicate the current margins used for SABR patients at the British Columbia Cancer Agency are sufficient in providing adequate tumor coverage when accounting for tumor motion and setup uncertainties. === Science, Faculty of === Physics and Astronomy, Department of === Graduate