Abstract
BACKGROUND AND PURPOSE: The accuracy of a surface-guided radiotherapy (SGRT) system for patient positioning in free-breathing breast cancer radiotherapy (RT) was evaluated against cone-beam computed tomography (CBCT)-based image registration. MATERIALS AND METHODS: A total of 106 treatment fractions from 20 patients who received breast RT in free-breathing in our institution were analysed. Patient positioning was performed using an SGRT system, without initial alignment to skin markers. The positioning deviations were reduced to within our clinical tolerance of ± 3.0 mm (translations) and ± 3.0° (rotations) before CBCT acquisition. CBCT-based setup corrections were calculated using bony anatomy registration between the CBCT and planning computed tomography (CT) images. The translational and rotational corrections derived from SGRT were compared with those from CBCT. A two-sample paired t-test with a significance level of 0.05 was performed. RESULTS: No significant differences were observed between SGRT and CBCT for longitudinal (p = 0.42) and lateral translations (p = 0.70). Vertical translation and rotations about all three axes were within clinical tolerance levels with SGRT, while CBCT-based corrections showed statistically significant differences (p < 0.05). CONCLUSIONS: The results support the feasibility of marker-free SGRT with CBCT verification as a reliable setup method in free-breathing breast cancer RT.