Affiliations: [a] State Key Laboratory of Oncology in South China, Sun Yat-sen University Cancer Center, Guangzhou, China
| [b] Department of Oncology, The Seventh Affiliated Hospital, Sun Yat-sen University, Shenzhen, China
Correspondence:
[*]
Corresponding authors. Feng Chi, State Key Laboratory of Oncology in South China, Sun Yat-sen University Cancer Center, 651 East Dongfeng Road, Guangzhou, China, 510060. E-mail: chifeng@sysucc.org.cn and Xiaofen Pan, Department of Oncology, The Seventh Affiliated Hospital, Sun Yat-sen University. No. 628, Zhenyuan Road, Guangming District, Shenzhen, 518100 China. E-mail: panxf8@mail.sysu.edu.cn.
Abstract: Background:There is no standardized registration frame of cone beam CT (CBCT) in intensity modulated radiotherapy (IMRT) for nasopharyngeal carcinoma (NPC). The overall registration frame that covers the whole head and neck is the most commonly used CBCT registration frame for NPC patients in IMRT. Objective:To compare the set-up errors using different registration frames of CBCT for NPC to assess the set-up errors for different region of the commonly used clinical overall registration frame. Methods:294 CBCT images of 59 NPC patients were collected. Four registration frames were used for matching. The set-up errors were obtained using an automatic matching algorithm and then compared. The expansion margin from the clinical target volume (CTV) to the planned target volume (PTV) in the four groups was also calculated. Results:The average range of the isocenter translation and rotation errors of four registration frames are 0.89∼2.41 mm and 0.49∼1.53°, respectively, which results in a significant difference in the set-up errors (p < 0.05). The set-up errors obtained from the overall frame are smaller than those obtained from the head, upper neck, and lower neck frames. The margin ranges of the overall, head, upper neck, and lower neck frames in three translation directions are 1.49∼2.39 mm, 1.92∼2.45 mm, 1.86∼3.54 mm and 3.02∼4.78 mm, respectively. The expansion margins calculated from the overall frame are not enough, especially for the lower neck. Conclusion:Set-up errors of the neck are underestimated by the overall registration frame. Thus, it is important to improve the position immobilization of the neck, especially the lower neck. The margin of the target volume of the head and neck region should be expanded separately if circumstances permit.
