Room: Exhibit Hall | Forum 5
Purpose: To explore feasibility of cone-beam CT (CBCT) simulation for planning same-day volumetric modulated arc therapy (VMAT) LS-spine treatments and to develop a method to compensate for the limited field-of-view (FOV) in CBCT simulation for patients with larger separation.
Methods: Dosimetric properties of seven LS-spine VMAT patient plans transferred onto their daily IGRT CBCTs were evaluated. An anthropomorphic pelvic phantom was used to acquire one central and two pairs of CBCTs with lateral couch shifts of +/-5 and +/-10 cm. The left/right CBCTs for each pair were stitched in MATLAB to create one large-FOV panoramic CBCT (pCBCT) image. The quality of the stitched images was compared to central CBCT and CT images, in terms of artifacts and noise levels. VMAT treatment plans created in Pinnacle on CT scan of the phantom were transferred onto pCBCT images and evaluated dosimetrically. The CBCT stitching method was also applied to one patient dataset with truncated left/right CBCTs and the VMAT plan was evaluated on the pCBCT image.
Results: PTV-D90 and maximum dose values for VMAT plans calculated on daily CBCTs for the seven patients were on average within 0.6% and 1.9% of original values, respectively, and for the patient plan on pCBCT within 0.4% and 1.1%. Minor artifacts were observed in lateral CBCTs and pCBCTs, worsening with larger shifts. Mean noise levels of CT, central CBCT, and pCBCT images were 1.1%, 1.5%, and 1.4%, respectively. PTV-D90 for VMAT plan transferred onto phantom pCBCTs with small and large shifts were, respectively, 0.1% and 0.2% smaller than original plan.
Conclusion: This study shows that planning palliative VMAT LS-spine treatments on CBCT images provides clinically acceptable dosimetric calculations. Large-FOV pCBCT with comparable image quality can be achieved using the method presented herein. However, target volume delineation on CBCT images as compared to CT needs further investigation.