摘要
神经影像生物标志物为脑血管病的诊断提供了依据,随着传统的人工诊疗模式转向信息化的医学诊疗模式,深度学习技术在脑血管病的辅助诊疗中展现了巨大的潜力。使用深度学习算法可对脑血管病影像中的深层特征进行提取和分析,并应用于临床的辅助诊断、预后评估等方面。本文介绍了深度学习模型的发展历程及评价指标,阐述了深度学习技术在脑血管病影像中的应用进展,包括血管结构和病变分割、疾病识别和预测判断,指出了深度学习在脑血管病辅助诊疗中面临的挑战,并对未来的发展前景进行了展望。
Neuroimaging biomarkers provide a basis for the diagnosis of cerebrovascular diseases.With the traditional manual diagnosis and treatment mode turning to the information-based one,deep learning technology shows great potential in the auxiliary diagnosis and treatment of cerebrovascular diseases.Deep learning algorithm can extract and analyze deep features in cerebrovascular disease images,and be applied to clinical auxiliary diagnosis,prognosis evaluation and other aspects.Firstly,the development process and evaluation indexes of deep learning model is introduced.Then,it focuses on the application progress of deep learning technology in cerebrovascular disease images,including vascular structure and lesion segmentation,disease recognition and prediction judgment.At last,the challenges of deep learning in the auxiliary diagnosis and treatment of cerebrovascular diseases and the future development prospect are discussed.
作者
徐佳薇
武杰
雷宇
XU Jiawei;WU Jie;LEI Yu(School of Health Science and Engineering,University of Shanghai for Science&Technology;Department of Neurosurgery,Huashan Hospital,Fudan University)
出处
《中国医学计算机成像杂志》
CSCD
北大核心
2022年第2期216-220,共5页
Chinese Computed Medical Imaging
基金
国家自然科学基金(61605114)。
关键词
脑血管病
深度学习
病灶分割
预测判断
Cerebrovascular disease
Deep learning
Lesion segmentation
Prediction judgment
