影像组学分类甲状腺结节良恶性的研究进展  

Research Progress of Radiomics in Classifying Benign and Malignant Thyroid Nodules

作　　者：汪永凤 欧小红 陈思元[2] 韦智晓[2] WANG Yongfeng;OU Xiaohong;CHEN Siyuan;WEI Zhixiao(Department of Medical Equipment,The First Affiliated Hospital of Guangxi Medical University,Nanning Guangxi 530000,China;Department of Nuclear Medicine,The First Affiliated Hospital of Guangxi Medical University,Nanning Guangxi 530000,China)

机构地区：[1]广西医科大学第一附属医院医疗设备科,广西南宁530000 [2]广西医科大学第一附属医院核医学科,广西南宁530000

出　　处：《中国医疗设备》2023年第7期170-176,共7页China Medical Devices

基　　金：广西壮族自治区卫生健康委员会自筹经费科研课题(2021-KY-桂卫)。

摘　　要：甲状腺结节是一种常见的甲状腺疾病。近年来,甲状腺癌发病率逐年升高,且受新型冠状病毒的影响,其死亡率也在进一步增加。影像组学是一种从医疗影像中提取高通量定量化特征并对其进行分析以达到疾病诊断或预后预测等目的的方法。传统影像组学是基于手工特征的分析,而基于深度学习的影像组学在自动化分析方面更具优势。本文对影像组学在甲状腺结节良恶性分类中的研究进展进行综述。Thyroid nodules are a common thyroid disease.In recent years,the morbidity of thyroid cancer has increased,and the mortality rate is also further increasing due to the influence of COVID-19.Radiomics is a method of extracting and analyzing highthroughput quantitative features from medical images to achieve the goal of disease diagnosis or prognosis prediction.The traditional radiomics is based on manual feature analysis,while deep learning-based radiomics has more advantages in automated analysis.This paper reviews the research progress of radiomics in classifying benign and malignant thyroid nodules.

关 键 词：甲状腺结节 甲状腺癌 影像组学 深度学习 

分 类 号：R581[医药卫生—内分泌]