Workflow and error analyses of patient setup based on open-face mask immobilization combined with AlignRT for head tumor radiotherapy
李俊禹,吴昊,杨敬贤,周舜,卢子红,于松茂,陈吉祥,王美娇,姚凯宁,杜乙
DOI: https://doi.org/10.3760/cma.j.cn112271-20220510-00199
2022-01-01
Abstract:Objective:To propose a markless patient setup workflow based on the optical surface monitoring system (AlignRT) and open-face mask immobilization for whole-course head tumor radiotherapy, assess the setup time and repositioning frequency of the proposed workflow, and conduct a comparative analysis of the differences, correlation, and consistency of the setup errors of the AlignRT and cone beam CT (CBCT) systems.Methods:A retrospective analysis was conducted for the data on the errors of 132 fractionated setup based on open-face mask immobilization of 33 head tumor patients. AlignRT-guided markless patient setup workflow was applied throughout the radiotherapy. Meanwhile, the body structures automatically generated by the treatment planning system were used as body references. The 6-degree-of-freedom (6DoF) setup errors (lateral, vertical, longitudinal, rotation, pitch, roll, and yaw directions), setup time, and repositioning frequency of the AlignRT and CBCT systems were recorded and analyzed. The Wilcoxon and Spearman analyses were used to statistically assess the differences and correlation of the setup errors of the two systems. Moreover, the Bland-Altman analysis was employed to evaluate the consistency of the two systems.Results:The 6DoF setup errors of CBCT were within the clinical tolerance (linear motions: -0.30 to 0.30 cm; rotational motions: -2.0° to 2.0°). The setup time and repositioning frequency of CBCT were (98 ± 31) s and 1.51% (2/132), respectively. There was no significant difference in setup errors between the two systems except those in x-axis ( Z = -3.11, P= 0.002), y-axis ( Z = -7.40, P<0.001), and Pitch ( Z= -4.48, P<0.001). There was a significant positive correlation between the setup errors along lateral ( rs = 0.47, P<0.001) and vertical ( rs = 0.29, P = 0.001) directions, rotation (Rtn; rs = 0.47, P<0.001), pitch (Pitch; rs = 0.28, P = 0.001) and roll (Roll; rs = 0.45, P<0.001) of the two systems. The 95% limits of agreement (95% LoA) of 6DoF setup errors were -0.12 to 0.09 cm, -0.07 to 0.17 cm, -0.19 to 0.20 cm, -1.0° to 0.9 °, -1.0° to 1.5°, and -0.9° to 1.0°, respectively. The 95% confidence interval (95% CI) of 95% LoA was -0.14 to 0.11 cm, -0.09 to 0.19 cm, -0.23 to 0.23 cm, -1.2° to 1.1°, -1.2° to 1.7°, and-1.0° to 1.1°, respectively, all of which were within the permissible error ranges. The 6DoF setup error difference of 3.41% (27/792< 5%) was beyond the 95% LoA. The maximum absolute differences of 6DoF setup errors within the 95% LoA were 0.12, 0.16, 0.19 cm, 0.9°, 1.5°, and 1.0°, respectively. Conclusions:The proposed markless setup workflow based on AlignRT combined with open-face mask immobilization for whole-course head tumor radiotherapy exhibits reasonable agreement and consistency with the patient setup using CBCT, with acceptable clinical efficiency. It can be applied to the first radiotherapy and the real-time monitoring of therapy to improve the safety and thus is of value in clinical applications.