基于深度学习的男性盆腔CT影像自动分割概述  

Overview of Automatic Segmentation of Male Pelvic CT Images Based on Deep Learning

作　　者：周瑾 刘国才[2] 鞠忠建 Zhou Jin;Liu Guocai;Ju Zhongjian(School of Physical Science and Technology,Wuhan University,Wuhan 430072,Hubei,China;College of Electrical and Information Engineering,Hunan University,Changsha 410082,Hunan,China;Department of Radiation Oncology,First Medical Center,Chinese PLA General Hospital,Beijing 100853,China)

机构地区：[1]武汉大学物理科学与技术学院,武汉430072 [2]湖南大学电气与信息工程学院,长沙410082 [3]中国人民解放军总医院第一医学中心放疗科,北京100853

出　　处：《肿瘤预防与治疗》2021年第1期6-15,共10页Journal of Cancer Control And Treatment

基　　金：国家自然科学基金(编号:61671204);解放军总医院临床科研扶持基金(编号:2017 FC-TSYS-3027)

摘　　要：近年来,深度学习已逐渐应用于放射治疗的器官自动分割和勾画。但是基于CT图像的盆腔器官自动分割仍具有较大挑战性。本文介绍了图像分割常用的基础网络模型和框架,以及适用于医学图像分割的网络、损失函数、常用数据集改进,重点概述了近五年基于CT图像使用深度学习自动分割男性盆腔器官的主要网络和结果,探讨了深度学习自动分割所面临的挑战和局限性,以及未来潜在的研究方向。In recent years,deep learning has been gradually applied to the automatic segmentation and delineation of organs in radiotherapy.However,automatic segmentation of pelvic organs based on CT images is still challenging.This paper introduces basic network models and frameworks commonly used in image segmentation,as well as the improvement of network,loss function,and common datasets which is suitable for medical image segmentation.It focuses on the main networks and results of deep learning in segmenting male pelvic organs automatically based on CT images in the past five years.Finally,challenges and limitations of deep learning in automatic segmentation are discussed,as well as the potential research directions in the future.

关 键 词：深度学习 医学图像 自动分割 CT图像 放射治疗 男性盆腔 

分 类 号：R73-33[医药卫生—肿瘤] TP389.1[医药卫生—临床医学]