宫颈癌后装治疗中基于U-net的自动施源器分割  

Automatic applicator segmentation based on U-net model in the brachytherapy of cervical cancer

作　　者：胡海 黎杰[2] 王培[2] 唐斌[2] 王先良[2] 杨强[1,2] HU Hai;LI Jie;WANG Pei;TANG Bin;WANG Xianliang;YANG Qiang(The Applied Nuclear Technology in the Geosciences Key Laboratory of Sichuan province,Chengdu University of Technology,Chengdu 610059,China;Sichuan Cancer Hospital & Institute)

机构地区：[1]成都理工大学地学核技术四川省重点实验室,成都610059 [2]四川省肿瘤医院·研究所

出　　处：《实用肿瘤学杂志》2021年第3期248-253,共6页Practical Oncology Journal

基　　金：四川省重点研发计划(编号:2019YFS0473),四川省卫生健康科研课题普及项目(编号:19PJ273)。

摘　　要：目的在CT引导的宫颈癌三维后装治疗计划制定中,应用U-net模型完成施源器的自动分割。方法基于U-net网络创建深度学习模型,将2019年12月—2020年10月的27例宫颈癌患者数据经过预处理后写入数据集,按照15∶2∶10的比例将数据集划分为训练集、验证集和测试集。将训练集和验证集数据放入模型中训练并验证,并将测试集数据应用到训练好的神经网络中分割出施源器,采用戴斯相似性系数(DSC)、95百分位豪斯多夫距离(HD95)、相关体积差异(RVD)、精确率和召回率对模型进行评价。结果10例测试集患者平均的DSC为0.90,HD95为1.26 mm,RVD为-0.06,精确率为0.94,召回率为0.88,分割时间为5 s。结论本研究利用U-net网络实现了宫颈癌三维后装治疗计划制定施源器的自动分割,可将其应用于施源器的重建,于实现临床计划制定的自动化具有较大意义。Objective In the preparation of CT-guided three-dimensional brachytherapy treatment plan for cervical cancer,the U-net model was used to complete the automatic segmentation of the applicator.Methods A deep learning model was created based on the U-net network.The data of 27 cervical cancer patients from December 2019 to October 2020 were preprocessed and written into the data set,which was divided into ratios of 15∶2∶10 for a training set,a verification set and a test set,respectively.The training set and verification set were put into the model for training and verification.The test set was applied to the trained neural network to segment the applicators.Dice similarity coefficient(DSC),95th percentile house dove distance(HD95),relative volume diffidence(RVD),precision and recall were used to evaluate this model.Results The average DSC of 10 patients in the test set was 0.90,HD95 was 1.26mm,RVD was-0.06,the accuracy rate was 0.94,the recall rate was 0.88,and the segmentation time was 5s.Conclusion In this study,the U-net network was used to realize the automatic segmentation of the applicator in the three-dimensional brachytherapy treatment plan for cervical cancer.It can be applied to the reconstruction of the applicator,which has great significance in realizing the automation of clinical planning.

关 键 词：深度学习 施源器分割 后装 宫颈癌 

分 类 号：R737.3[医药卫生—肿瘤]