MR引导食管癌自适应放射治疗研究进展  

Advances in MR-guided adaptive radiotherapy for esophageal cancer

作　　者：王华东 李昊东 李振江 尹勇 WANG Huadong;LI Haodong;LI Zhenjiang;YIN Yong(Department of Graduate,Shandong First Medical University and Shandong Academy of Medical Sciences,Jinan250117,China;Department of Radiation Physics Technology,Shandong Cancer Hospital and Institute,Shandong First Medical University and Shandong Academy of Medical Science,Jinan 250117,China)

机构地区：[1]山东第一医科大学(山东省医学科学院)研究生部,山东济南250117 [2]山东省肿瘤防治研究院(山东省肿瘤医院)放射物理技术科,山东第一医科大学(山东省医学科学院),山东济南250117

出　　处：《中华肿瘤防治杂志》2023年第5期301-307,共7页Chinese Journal of Cancer Prevention and Treatment

基　　金：国家自然科学基金(82102173,82072094);山东省自然科学基金(ZR2019LZL017);2021山东省医学会临床研究基金-齐鲁专项(YX H2022ZX02198)。

摘　　要：目的对MR引导下食管癌自适应放射治疗的潜在优势、面临的问题以及与深度学习相结合的发展趋势进行综述。方法以“MR、食管癌、自适应放射治疗、自动勾画”为中文关键词,“MR、esophageal cancer、adaptive radiotherapy、automatic delineation”为英文关键词,系统检索中国知网及PubMed数据库2010-01-01-2022-08-30相关文献。纳入标准:(1)食管癌自适应放射治疗相关文献;(2)MR引导的自适应放射治疗相关文献;(3)在自适应放射治疗过程中应用人工智能自动勾画相关文献。排除标准:数据陈旧文献。最终共纳入62篇文献。结果集成的MR引导放射治疗系统(例如MR-Linac)可以通过使用实时成像和运动追踪来克服由于呼吸引起的食管运动。另外,MR-Linac通过靶区轮廓的精确识别、呼吸门控以及在线自适应计划可以使投照剂量更加严格和准确,保证靶区足量的同时又能减少周围危及器官的受照射剂量,从而使放疗引起的心脏和肺的毒副作用减少。最后,随着深度学习人工智能在图像分割和三维剂量预测领域的快速发展,实现了靶区的快速勾画和治疗计划的快速生成,为短时间内实现MR引导的在线自适应放射治疗提供了技术支持。结论MR引导的自适应放射治疗投照剂量更加严格和准确,未来需要进一步结合人工智能,实现更快地自动勾画与自动计划,缩短治疗时间,减小靶区和危及器官发生形变与位移的概率,进一步提升放射治疗的准确性。Objective To review the potential advantages,problems and development trend of MR guided adaptive radiotherapy for esophageal cancer.Methods With"MR,esophageal cancer,adaptive radiotherapy,automatic delineation"as the keywords,systematically searched the relevant literatures of CNKI and PubMed database from January 1,2010 to August 30,2022.Inclusive criteria:(1)literatures on adaptive radiotherapy for esophageal cancer.(2)Literatures on MR guided adaptive radiotherapy.(3)In the process of adaptive radiotherapy,artificial intelligence was applied to automatically sketch relevant literatures.Exclusion criteria:outdated data.Finally,62 articles were included.Results Integrated MR guided radiotherapy systems(such as MR Linac)can overcome esophageal movements due to respiration by using real-time imaging and motion tracking.In addition,MR Linac can make the radiation dose more strict and accurate through accurate identification of target area contour,respiratory gating and online adaptive planning,ensure sufficient target area and reduce the radiation dose of surrounding organs at the same time,thus reducing the side effects of heart and lung caused by radiotherapy.Finally,with the rapid development of deep learning artificial intelligence in the field of image segmentation and three-dimensional dose prediction,the rapid delineation of target areas and the rapid generation of treatment plans have been achieved,providing technical support for MR guided online adaptive radiotherapy in a short time.Conclusion The projection dose of MR guided adaptive radiotherapy is more strict and accurate.In the future,it is necessary to further combine artificial intelligence to achieve faster automatic sketching and automatic planning,shorten treatment time,reduce the probability of deformation and displacement of target area and endangered organs,and further improve the accuracy of radiotherapy.

关 键 词：磁共振成像 食管癌 自适应放射治疗 自动勾画 综述文献 

分 类 号：R735.1[医药卫生—肿瘤] R730.55[医药卫生—临床医学]