卷积神经网络在轻度认知障碍中的影像学研究进展  

Imaging research progress in mild cognitive impairment using convolutional neural networks

作　　者：李晓陵[1] 王敬贤 李昂[3] 李孟[1] 曹丹娜[1] 王丰[1] 崔璇 姚春丽 蔡丽娜 LI Xiaoling;WANG Jingxian;LI Ang;LI Meng;CAO Danna;WANG Feng;CUI Xuan;YAO Chunli;CAI Lina(First Hospital Affiliated to Heilongjiang University of Chinese Medicine,Harbin 150040,China;Graduate School of Heilongjiang University of Chinese Medicine,Harbin 150040,China;Public Health Institute of Harbin Medical University,Harbin 150081,China)

机构地区：[1]黑龙江中医药大学附属第一医院,哈尔滨150040 [2]黑龙江中医药大学研究生院,哈尔滨150040 [3]哈尔滨医科大学公共卫生学院,哈尔滨150081

出　　处：《磁共振成像》2021年第9期88-90,94,共4页Chinese Journal of Magnetic Resonance Imaging

基　　金：国家自然科学基金面上项目(编号:82074537、81973930、81373714);黑龙江省自然科学基金项目(编号:LH2020H103、H2016081);哈尔滨市科技创新人才优秀学科带头人基金(编号:2016RAXYJ096);教育部“春晖计划”(编号:Z2009-1-15030);哈尔滨市科技创新人才专项资金项目(编号:2017RAQXJ180);黑龙江省中医药科研项目(编号:ZHY12-Z046);黑龙江省卫生厅科研课题(编号:2012-346)。

摘　　要：人类连接组计划是脑成像领域最高级别的研究项目,人工智能(artificial intelligence,AI)是脑科学研究过程中必不可少的工具。卷积神经网络(convolutional neural networks,CNN)作为AI领域的最新技术之一,在计算机视觉、影像处理等方面表现卓著,于轻度认知障碍(mild cognitive impairment,MCI)的诊断和分析等方面展示出巨大的临床应用前景。作者以卷积神经网络为线索,探讨其在MCI领域的研究现状与未来发展方向。The Human Connected Group Project is the highest-level research project in the field of brain imaging,and artificial intelligence(AI)is an indispensable tool in the process of brain science research.As one of the latest technologies in the field of AI,convolutional neural networks(CNN)have outstanding performance in computer vision,image processing,etc.,and have shown great clinical application prospects in the diagnosis and analysis of mild cognitive impairment(MCI).The author uses CNN as a clue to discuss its current research status and future development direction in the field of MCI.

关 键 词：卷积神经网络 轻度认知障碍 脑 功能磁共振成像 结构磁共振成像 

分 类 号：R445.2[医药卫生—影像医学与核医学] R749.1[医药卫生—诊断学]