摘要
目的自主研发一款基于深度学习的放疗轮廓勾画软件(Yorktal-CS),旨在提高轮廓勾画的效率,提升现有轮廓勾画软件的水平。方法使用C++语言研发软件基本架构,采用基于深度学习神经网络技术研发轮廓自动勾画功能,并对软件进行验证与确认。结果完成了放疗轮廓勾画软件研发,修复软件缺陷16个,并获得软件实物一套,软件性能得到提升。结论该软件具有便捷、快速、高效等优势,可提高临床放疗轮廓勾画效率,在一定程度上促进放疗同质化发展。
Objective To research and develop a deep learning-based radiotherapy contouring software(Yorktal-CS)independently,which aims to improve the efficiency of contouring and improve the level of existing contouring software.Methods The basic architecture of the software was developed using C++language.Based on the deep learning neural network technology,the function of automatic contouring was developed and the software was verified and confirmed.Results We completed the development of the radiotherapy contouring software,and repaired 16 software bugs.A set of software objects was obtained,which improved the performance of the software.Conclusion This software has the advantages of convenience,rapidity and high efficiency,which can improve the contouring efficiency of clinical radiotherapy and promote the development of homogeneous radiotherapy to a certain extent.
作者
罗邦雨
伍亚军
周一兵
钱金栋
赵利荣
崔天祥
钟良志
李光辉
孙建国
LUO Bangyu;WU Yajun;ZHOU Yibing;QIAN Jindong;ZHAO Lirong;CUI Tianxiang;ZHONG Liangzhi;LI Guanghui;SUN Jianguo(Cancer Institute,The Second Affiliated Hospital of The Army Medical University,Chongqing 400037,China;Yorktal Digital Medical Imaging Technology Co.,Ltd,Shenzhen Guangdong 518048,China)
出处
《中国医疗设备》
2020年第7期40-44,共5页
China Medical Devices
基金
国家重点研发计划项目(2016YFC0106400)
重庆市技术创新与应用发展专项面上项目(cstc2019jscxmsxmX0265)。
关键词
放射治疗
自动轮廓勾画
深度学习
危及器官
同质化放疗
radiotherapy
automatic delineation
deep learning
organs at risk
homogeneous radiotherapy
