摘要目的:比较国产iSCOUT图像引导定位系统和Varian加速器机载电子射野成像装置（EPID）位置验证系统的体位验证准确性。方法:回顾性分析中国医学科学院肿瘤医院放疗科2020年11月到2023年11月期间收治的573例在调强放射治疗过程中同时使用2种定位系统进行体位验证的患者，其中头颈肩架固定的头颈肿瘤47例，乳腺托架固定的乳腺癌75例，胸腹平架固定的宫颈癌147例，腹盆架固定的直肠癌304例。放疗第1周至少3次体位验证，之后每周至少1次体位验证。比较患者同一体位2种系统的摆位误差和一致性，对误差数据进行Wilcoxon符号秩检验及kappa一致性分析。结果:使用iSCOUT和EPID两种系统的摆位误差如下：头颈肿瘤 X轴方向0.10（0.00，0.20）cm和0.10（0.00，0.20）cm， P<0.001；乳腺癌 Z轴方向0.10（-0.10，0.30）cm和0.00（-0.20，0.20）cm， P=0.030；直肠癌 X轴方向0.00（-0.20，0.10）cm和0.00（-0.10，0.20）cm， P<0.001， Y轴方向-0.10（-0.30，0.10）cm和-0.10（-0.30，0.20）cm， P=0.009，均差异有统计学意义，其余肿瘤类型其他方向，2种系统的差别无统计学意义。所有肿瘤类型的3个方向2种系统所获取的误差数据kappa一致性检验κ值结果均>0.5，表明2种系统的一致性均较强。在摆位误差≤2 mm区间内，头颈肿瘤iSCOUT组在 X轴（89.80%比82.99%）、 Y轴（90.82%比90.14%）及 Z轴（92.86%比89.46%）表现更优；其余部位肿瘤在>4 mm的较大误差区间，EPID组总体表现更稳定。 结论:图像引导调强放疗中，国产iSCOUT图像引导定位系统表现出非劣于Varian加速器机载EPID位置验证系统的效果，可作为临床实践中一种可行的选择方案。

abstractsObjective:To compare the setup verification accuracy of the domestic iSCOUT image-guided positioning system and the Varian accelerator airborne electronic portal imaging device (EPID) positioning verification system.Methods:A retrospective analysis was conducted of 573 patients treated in the Department of Radiation Oncology at the Cancer Hospital of the Chinese Academy of Medical Sciences between November 2020 and November 2023. All patients underwent position verification using both systems during intensity-modulated radiotherapy. The cohort included 47 head and neck cancer patients immobilized with thermoplastic masks, 75 breast cancer patients immobilized with breast brackets, 147 cervical cancer patients immobilized with thoracoabdominal flat devices, and 304 rectal cancer patients immobilized with abdominopelvic frames. At least 3 position verifications were performed during the first week of radiotherapy, followed by at least one verification per week thereafter. Setup errors and consistency between the two systems were compared for the same patient positions, and the error data were analyzed using the Wilcoxon signed-rank test and kappa consistency analysis.Results:The setup errors obtained using the iSCOUT and EPID systems were as follows: for head and neck tumors, in the X-axis direction, 0.10 (0.00, 0.20) cm and 0.10 (0.00, 0.20) cm, respectively ( P<0.001); for breast cancer, in the Z-axis direction, 0.10 (-0.10, 0.30) cm and 0.00 (-0.20, 0.20) cm, respectively ( P=0.030); for rectal cancer, in the X-axis direction, 0.00 (-0.20, 0.10) cm and 0.00 (-0.10, 0.20) cm, respectively ( P<0.001), and in the Y-axis direction, -0.10 (-0.30, 0.10) cm and -0.10 (-0.30, 0.20) cm, respectively ( P=0.009). These differences were statistically significant, whereas no significant differences were observed in other directions or tumor sites. The kappa consistency test showed κ values greater than 0.5 in all three directions for all tumor sites, indicating strong agreement between the two systems. Within the setup error interval of ≤2 mm, the iSCOUT group demonstrated better performance for head and neck tumors in the X-axis (89.80% vs. 82.99%), Y-axis (90.82% vs. 90.14%), and Z-axis (92.86% vs. 89.46%). For tumors at other sites, the EPID group showed overall more stable performance in the larger error interval of >4 mm. Conclusions:In image-guided intensity-modulated radiotherapy, the domestic iSCOUT image-guided positioning system demonstrated non-inferior performance compared with the Varian accelerator on-board EPID positioning verification system, and may serve as a feasible alternative in clinical practice.