Surface imaging, portal imaging, and skin marker set-up vs. CBCT for radiotherapy of the thorax and pelvis

Aim: The aim of this study was to compare surface imaging, portal imaging, and skin marker set-up in radiotherapy of thoracic and pelvic regions, using cone beam computed tomography (CBCT) data as the gold standard. Patients and methods: Twenty patients were included in this study. CBCT, surface acquisition (SA), and two orthogonal portal images (PI) were acquired during the first four treatment sessions. Patient set-up corrections, obtained by registering the planning CT with CBCT, were used as the gold standard. Registration results of the PI and SA were evaluated and compared with those obtained with CBCT. The advantage derived from using SA or PI verification systems over a skin marker set-up was also quantified. Results: A statistically significant difference between PI and SA (in favour of PI) was observed in seven patients undergoing treatment of the pelvic region and in two patients undergoing treatment of the thoracic region. The use of SA or PI, compared with a skin marker set-up, improved patient positioning in 50% and 57 % of the thoracic fractions, respectively. For pelvic fractions, the use of PI was beneficial in 73 % of the cases, while the use of SA was beneficial in only 45 %. Patient positioning worsened with SA, particularly along longitudinal and vertical directions. Conclusion: PI yielded more accurate registration results than SA for both pelvic and thoracic fractions. Compared with the skin marker set-up, PI performances were superior to SA for pelvic fractions while comparable results were obtained for thoracic fractions.