摘要
随着人工智能技术的发展,图像分割技术在直肠癌放疗领域有越来越多的研究和应用。新辅助放化疗是局部晚期直肠癌患者的标准治疗策略。对于放疗科医师而言,手动勾画放疗靶区及危及器官是一项非常费时耗力的工作。采用人工智能技术可以构建放疗靶区自动勾画模型,显著提高放疗靶区勾画的效率和鲁棒性。此外,结合影像组学方法,基于计算机断层扫描、磁共振成像等影像图像,提取直肠肿瘤区域的特征可以建立直肠癌新辅助治疗的疗效评估和预测模型,以帮助临床医师制订个体化的治疗方案。其中,分割感兴趣区,并且从中提取影像特征是模型构建的关键步骤。本文基于图像分割在直肠癌放疗领域中的应用展开综述,以探究图像分割对直肠癌放疗的重要性以及未来的研究方向。
With the rapid development of artificial intelligence technology,the research and application of image segmentation technology in the field of radiotherapy for rectal cancer have captivated increasing attention.Neoadjuvant chemoradiotherapy followed by radical surgery is the standard treatment for patients with locally advanced rectal cancer.Manual delineation of radiotherapy targets and organs at risk is a time-consuming and laborious task.Developing an automatic delineation model of radiotherapy targets using artificial intelligence can significantly improve the efficiency and robustness of target delineation.In addition,combined with radiomics methods,based on computed tomography(CT),magnetic resonance imaging(MRI),extracting radiation features from rectal tumor can build a model for efficacy evaluation and prediction of neoadjuvant therapy,which can help clinicians formulate individualized treatment regimens.Segmenting the region of interest(ROI)and extracting radiation features is a key step in model construction.This article will review the application of image segmentation in the field of radiotherapy for rectal cancer,aiming to explore the importance of image segmentation in radiotherapy for rectal cancer and future research directions.
作者
唐瑗玲
王辛
Tang Yuanling;Wang Xin(Department of Abdominal Cancer,West China Hospital,Sichuan University,Chengdu 610041,China;Department of Radiation Oncology,West China Hospital,Sichuan University,Chengdu 610041,China)
出处
《中华放射肿瘤学杂志》
CSCD
北大核心
2024年第9期859-863,共5页
Chinese Journal of Radiation Oncology
基金
国家自然科学基金面上项目(82073338)
四川大学华西医院学科卓越发展1·3·5工程临床研究孵化项目(2020HXFH002)。
关键词
直肠肿瘤
放射疗法
人工智能
图像分割
疗效预测
Rectalneoplasms
Radiotherapy
Artificial intelligence
Image segmentation
Efficacyprediction
