深度学习探讨视网膜成像在阿尔茨海默病管理中的应用  

Application of deep learning-based retinal imaging in management of Alzheimer’s disease

作　　者：吴美初 柯殷雨 唐云华 张富文 WU Meichu;KE Yinyu;TANG Yunhua;ZHANG Fuwen(Eye School of Chengdu University of Traditional Chinese Medicine,Chengdu 610075,Sichuan Province,China;Ineye Hospital,Chengdu University of Traditional Chinese Medicine,Chengdu 610084,Sichuan Province,China;Key Laboratory of Sichuan Province Ophthalmopathy Prevention&Cure and Visual Function Protection,Chengdu 610084,Sichuan Province,China)

机构地区：[1]成都中医药大学眼科学院,四川省成都市610075 [2]成都中医大银海眼科医院,四川省成都市610084 [3]中医药眼病防治与视功能保护四川省重点实验室,四川省成都市610075

出　　处：《眼科新进展》2024年第8期663-667,共5页Recent Advances in Ophthalmology

基　　金：四川省自然科学基金项目(编号:2023NSFSC0690)。

摘　　要：阿尔茨海默病(AD)是一种不可逆的神经系统退行性疾病,早期筛查和诊断对有效治疗AD至关重要,然而AD诊断的复杂性使得基层筛查面临困难。视网膜作为中枢神经系统的外延,其变化与AD进展紧密相连,能够为非入侵性早期检测AD提供新途径。近年来,深度学习(DL)在医学领域取得了显著进展,特别是图像识别和分析方面,DL与视网膜成像的结合在诊治AD中展现出巨大潜力。本文探讨了DL在分析AD视网膜图像中的应用进展,包括疾病的诊断、进展预测、长期管理策略,以及现阶段在临床实践中的不足,以期为DL在AD视网膜图像应用的进一步研究提供参考。Alzheimer’s disease(AD)is an irreversible neurodegenerative disorder,making early screening and diagnosis pivotal for its effective treatment.The complexity of diagnosing AD,however,poses significant challenges for primary-level screening.As an extension of the central nervous system,the retina exhibits changes closely linked to AD,offering a new pathway for non-invasive early detection of AD.In recent years,deep learning(DL)has achieved notable advancements in the medical field,especially in image recognition and analysis.The synergy between DL and retinal imaging has unveiled substantial potential in diagnosing and treating AD.This article reviews the advancements in applying DL to analyze retinal images related to AD,encompassing the diagnosis,prediction of progression,and long-term management of AD,as well as the current shortcomings in clinical practice,providing a reference for further research into the application of DL in retinal imaging for AD.

关 键 词：阿尔茨海默病 人工智能 深度学习 图像分类与分析 光学相干断层扫描 光学相干断层扫描血管成像 眼底照相 

分 类 号：R774[医药卫生—眼科]