摘要
介绍了甲状腺结节的传统诊断方法及局限性,阐述了CT纹理分析技术的原理及具体流程,综述了CT纹理分析技术在甲状腺结节筛查、甲状腺结节良恶性鉴别、预测甲状腺癌淋巴结转移及甲状腺结节关联基因表达中的研究进展,分析了CT纹理分析技术在甲状腺结节影像研究中存在的问题。指出了CT图像纹理特征提取的标准化、规范化以及结合深度学习算法的应用等是该领域未来的发展趋势。
The traditional diagnostic methods and limitations of thyroid nodules were introduced,and the principles and process of CT texture analysis technology were described.The research achievements and progress based on CT texture analysis in thyroid nodule screening,benign and malignant identification,prediction of lymph node metastasis and the correlation of gene expression were reviewed and analyzed,and the problems during the researches were explored.It's pointed out the future development in the field involved in the standardization and normalization of CT image texture feature extraction and the application of combined deep learning algorithms.
作者
茅枭骁
征锦
MAO Xiao-xiao;ZHENG Jin(Department of Imaging,Jingjiang People's Hospital of Medical College of Yangzhou University,Jingjiang 214500,Jiangsu Province,China;Department of Imaging,Northern Jiangsu People's Hospital of Yangzhou University,Yangzhou 225000,Jiangsu Province,China)
出处
《医疗卫生装备》
CAS
2020年第12期97-100,104,共5页
Chinese Medical Equipment Journal
