宫颈癌后装治疗中基于U-net的自动施源器分割

doi:10.11904/j.issn.1002-3070.2021.03.010

实用肿瘤学杂志 ›› 2021, Vol. 35 ›› Issue (3): 248-253.doi: 10.11904/j.issn.1002-3070.2021.03.010

宫颈癌后装治疗中基于U-net的自动施源器分割

胡海^1,2, 黎杰², 王培², 唐斌², 王先良², 杨强^1,2

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

收稿日期:2020-12-08 修回日期:2021-01-25 出版日期:2021-06-28 发布日期:2021-06-22
通讯作者: 杨强,E-mail:yq.mail@foxmail.com
作者简介:胡海,男,(1995-),硕士研究生,从事深度学习在放疗中应用的研究
基金资助:
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)

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

HU Hai^1,2, LI Jie², WANG Pei², TANG Bin², WANG Xianliang², YANG Qiang^1,2

1. The Applied Nuclear Technology in the Geosciences Key Laboratory of Sichuan province,Chengdu University of Technology,Chengdu 610059,China;
2. Sichuan Cancer Hospital & Institute

Received:2020-12-08 Revised:2021-01-25 Online:2021-06-28 Published:2021-06-22

摘要/Abstract

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。

关键词: 深度学习, 施源器分割, 后装, 宫颈癌

Abstract: 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.

Key words: Deep learning, Applicator segmentation, Brachytherapy, Cervical cancer

中图分类号:

R737.3

胡海, 黎杰, 王培, 唐斌, 王先良, 杨强. 宫颈癌后装治疗中基于U-net的自动施源器分割[J]. 实用肿瘤学杂志, 2021, 35(3): 248-253.

HU Hai, LI Jie, WANG Pei, TANG Bin, WANG Xianliang, YANG Qiang. Automatic applicator segmentation based on U-net model in the brachytherapy of cervical cancer[J]. Journal of Practical Oncology, 2021, 35(3): 248-253.

0
/ / 推荐

导出引用管理器 EndNote|Reference Manager|ProCite|BibTeX|RefWorks

链接本文: http://journal09.magtechjournal.com/Jwk3_syzlxzz/CN/10.11904/j.issn.1002-3070.2021.03.010

http://journal09.magtechjournal.com/Jwk3_syzlxzz/CN/Y2021/V35/I3/248

参考文献

1 Bray F,Ferlay J,Soerjomataram I,et al.Global cancer statistics 2018:GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries[J].CA Cancer J Clin,2018,68(6):394-424.
2 杨东明,卢庆刚."半野"三维适形放疗、固定野静态调强放疗和容积旋转调强放疗在中晚期宫颈癌治疗上的剂量学比较[J].实用肿瘤学杂志,2020,34(3):244-248.
3 于德洋,杨姗姗,云惟康,等.宫颈癌术后容积弧形调强与固定野动态调强放疗技术的剂量学比较[J].实用肿瘤学杂志,2017,31(2):152-155.
4 王先良,袁珂,康胜伟,等.环形施源器重建偏差对宫颈癌患者剂量影响分析[J].中华放射肿瘤学杂志,2017,26(3):316-319.
5 王先良,吴俊翔,袁珂,等.驻留点权重标准差与施源器位置不确定性对宫颈癌患者剂量影响分析[J].中华放射肿瘤学杂志,2017,26(4):419-422.
6 彭倩,王培,王先良,等.宫颈癌内外照射中基于CBCT外照射摆位误差分析[J].中华放射肿瘤学杂志,2016,25(10):1113-1116.
7 Ronneberger O,Fischer P,Brox T.U-Net:convolutional networks for biomedical image segmentation[C].Switzerland:Springer,2015.
8 Kingma DP,Ba J.Adam:a method for stochastic optimization[C].ICLR:San Diego,2015.
9 Zaffino P,Pernelle G,Mastmeyer A,et al.Fully automatic catheter segmentation in MRI with 3D convolutional neural networks:application to MRI-guided gynecologic brachytherapy[J].Phys Med Biol,2019,64(16):165008.
10 Taha AA,Hanbury A.Metrics for evaluating 3D medical image segmentation:analysis,selection,and tool[J].BMC Med Imaging,2015,15:29.
11 Chang HH,Zhuang AH,Valentino DJ,et al.Performance measure characterization for evaluating neuroimage segmentation algorithms[J].NeuroImage,2009,47(1):122-135.
12 EleKta.Brachytherapy applicators and accessories[M].Stockholm:Elekta AB,2017.
13 张凯文.基于U-Net的肝脏CT图像分割研究[D].广州:华南理工大学,2019.
14 朱琳,王学涛,戴振晖,等.CT图像快速施源器重建技术在三维后装放射治疗中的应用[J].临床医学工程,2014,21(6):683-684.
15 Zhang YP,Lei Y,Qiu R,et al.Multi-needle localization with attention U-Net in US-guided HDR prostate brachytherapy[J].Med Phys,2020,47(7):2735-2745.
16 谭燕,王培,郎锦义,等.放疗靶区勾画系统的设计与开发[J].中国医疗设备,2018,33(10):132-133.
17 Wang XL,Wang P,Li CR,et al.An automated dose verification software for brachytherapy[J].J Contemp Brachytherapy,2018,10(5):478-482.
18 王先良,王培,李厨荣,等.三维后装中一种逆向剂量优化算法[J].中华放射肿瘤学杂志,2020,29(8):676-681.
19 Wang XL,Wang P,Tang B,et al.An inverse dose optimization algorithm for three-dimensional brachytherapy[J].Front Oncol,2020,10:564-580.

[1]	姚丽萍, 王宏宗, 丁高恒, 刘玉琴. 2009—2015年甘肃省肿瘤登记地区宫颈癌流行特征及变化趋势分析[J]. 实用肿瘤学杂志, 2021, 35(3): 224-230.
[2]	秦伟, 何义富, 常娜, 钱立庭, 黄大兵, 崔亚云. 局部晚期宫颈癌放射治疗对直肠功能影响的研究进展[J]. 实用肿瘤学杂志, 2021, 35(2): 170-173.
[3]	任思颖,秦明芳,陈杨,石青萍,成会荣,高娇,文洪梅. 云南省肿瘤登记地区2011—2015年宫颈癌发病和死亡特征及时间趋势[J]. 实用肿瘤学杂志, 2021, 35(1): 13-17.
[4]	李静, 王玉霞, 陈杰. 长链非编码RNA HOTAIR在宫颈癌中的相关研究进展[J]. 实用肿瘤学杂志, 2020, 34(4): 368-371.
[5]	罗丽芳, 刘运铎, 关玥, 陈秀玮. 基于槲皮素标记的金属有机框架放疗联合阿霉素诱导宫颈癌HeLa细胞凋亡的研究[J]. 实用肿瘤学杂志, 2020, 34(3): 203-207.
[6]	秦凯, 程熠, 袁响林, 简丹妮, 张菁, 黄乐. 超声膀胱容量监测技术在宫颈癌术后调强放疗中的应用价值[J]. 实用肿瘤学杂志, 2020, 34(3): 232-236.
[7]	杨东明, 卢庆刚. “半野”三维适形放疗、固定野静态调强放疗和容积旋转调强放疗在中晚期宫颈癌治疗上的剂量学比较[J]. 实用肿瘤学杂志, 2020, 34(3): 244-248.
[8]	曹宁, 夏百荣, 娄阁. HPV蛋白、miRNAs和外泌体在宫颈癌炎症作用中的研究进展[J]. 实用肿瘤学杂志, 2020, 34(3): 282-285.
[9]	周瑶瑶, 陈凤云, 屠文骆. 棕榈酸在体外通过促进凋亡作用抑制宫颈癌Hela细胞增殖的研究[J]. 实用肿瘤学杂志, 2019, 33(5): 402-406.
[10]	申利, 魏军成. miRNA-409-3p表达水平对宫颈癌细胞增殖及顺铂化疗敏感性的影响[J]. 实用肿瘤学杂志, 2019, 33(4): 300-304.
[11]	李景, 吴华珍, 刘继琐. miR-362靶向Six1抑制宫颈癌细胞增殖、迁移的作用机制研究[J]. 实用肿瘤学杂志, 2019, 33(4): 310-316.
[12]	梁虹, 李奇, 王耀先, 靳红, 王铜, 王亚楠, 段雅妮, 宋敏. 宫颈癌化疗心血管毒性预防的间断时间序列试验[J]. 实用肿瘤学杂志, 2019, 33(2): 134-138.
[13]	成海燕, 马晶, 冯紫微, 黄可金, 王玉霞, 张云艳. 抑癌基因PRDM5在HPV16病毒致宫颈鳞癌发生过程中的作用研究[J]. 实用肿瘤学杂志, 2018, 32(6): 483-487.
[14]	付艳丽, 余杨, 卫玮, 邹晶晶, 孙翔. 高危型人乳头状瘤病毒载量和Th1/Th2不均衡表达预测宫颈癌变进程的研究[J]. 实用肿瘤学杂志, 2018, 32(6): 533-537.
[15]	秦文艳综述, 卢艳审校. 宫颈癌前哨淋巴结微转移的研究进展[J]. 实用肿瘤学杂志, 2018, 32(6): 571-574.

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

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价