Liu Zhikai, Liu Fangjie, Chen Wanqi, Tao Yinjie, Liu Xia, Zhang Fuquan, Shen Jing, Guan Hui, Zhen Hongnan, Wang Shaobin, Chen Qi, Chen Yu, Hou Xiaorong
Department of Radiation Oncology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences, Beijing, 100730, People's Republic of China.
Department of Radiation Oncology, Sun Yat-sen University Cancer Center, State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Guangzhou, 510060, People's Republic of China.
Cancer Manag Res. 2021 Nov 2;13:8209-8217. doi: 10.2147/CMAR.S330249. eCollection 2021.
Delineation of clinical target volume (CTV) and organs at risk (OARs) is important for radiotherapy but is time-consuming. We trained and evaluated a U-ResNet model to provide fast and consistent auto-segmentation.
We collected 160 patients' CT scans with breast cancer who underwent breast-conserving surgery (BCS) and were treated with radiotherapy. CTV and OARs were delineated manually and were used for model training. The dice similarity coefficient (DSC) and 95th percentile Hausdorff distance (95HD) were used to assess the performance of our model. CTV and OARs were randomly selected as ground truth (GT) masks, and artificial intelligence (AI) masks were generated by the proposed model. Two clinicians randomly compared CTV score differences of the contour. The consistency between two clinicians was tested. Time cost for auto-delineation was evaluated.
The mean DSC values of the proposed method were 0.94, 0.95, 0.94, 0.96, 0.96 and 0.93 for breast CTV, contralateral breast, heart, right lung, left lung and spinal cord, respectively. The mean 95HD values were 4.31mm, 3.59mm, 4.86mm, 3.18mm, 2.79mm and 4.37mm for the above structures, respectively. The average CTV scores for AI and GT were 2.89 versus 2.92 when evaluated by oncologist A (=0.612), and 2.75 versus 2.83 by oncologist B (=0.213), with no statistically significant differences. The consistency between two clinicians was poor (kappa=0.282). The time for auto-segmentation of CTV and OARs was 10.03 s.
Our proposed model (U-ResNet) can improve the efficiency and accuracy of delineation compared with U-Net, performing equally well with the segmentation generated by oncologists.
勾画临床靶区（CTV）和危及器官（OARs）对放射治疗很重要，但耗时较长。我们训练并评估了一个U-ResNet模型，以提供快速且一致的自动分割。
我们收集了160例接受保乳手术（BCS）并接受放射治疗的乳腺癌患者的CT扫描图像。手动勾画CTV和OARs，并用于模型训练。采用骰子相似系数（DSC）和第95百分位数豪斯多夫距离（95HD）来评估我们模型的性能。随机选择CTV和OARs作为真实（GT）掩码，并由所提出的模型生成人工智能（AI）掩码。两名临床医生随机比较轮廓的CTV评分差异。测试了两名临床医生之间的一致性。评估了自动勾画的时间成本。
所提出方法的平均DSC值分别为：乳腺CTV为0.94、对侧乳腺为0.95、心脏为0.94、右肺为0.96、左肺为0.96、脊髓为0.93。上述结构的平均95HD值分别为4.31mm、3.59mm、4.86mm、3.18mm、2.79mm和4.37mm。肿瘤学家A评估时，AI和GT的平均CTV评分为2.89对2.92（=0.612），肿瘤学家B评估时为2.75对2.83（=0.213），无统计学显著差异。两名临床医生之间的一致性较差（kappa=0.282）。CTV和OARs的自动分割时间为10.03秒。
与U-Net相比，我们提出的模型（U-ResNet）可以提高勾画的效率和准确性，与肿瘤学家生成的分割效果相当。
