[关键词]
[摘要]
随着人工智能技术的发展和普及，医学领域也出现了越来越多人工智能(AI)的身影。人工神经网络等新技术与临床的结合正成为研究热点，其中卷积神经网络(CNN)的深度学习算法在图像识别领域取得了巨大的成就，逐渐被用于糖尿病视网膜病变(DR)、年龄相关性黄斑变性(ARMD)、早产儿视网膜病变(ROP)、青光眼和白内障等多种眼科疾病的诊断和筛查中。目前针对不同眼科疾病，世界范围已有多个公开数据库，包括了眼底彩照、光学相干断层扫描(OCT)等多种图像资料，为眼科领域深度学习算法的训练和构建奠定了基础。同时算法本身也在不断优化，使相关AI产品的构建朝着更简便高效的方向发展，同时其临床运用也面临医学伦理和准入标准的问题。总之，深度学习算法的使用为几种常见眼科疾病的筛查诊断带来了巨大的改变也带来挑战，目前尚未大规模的投入临床应用中。本文针对人工智能在眼部疾病中的应用进展做综述，旨在总结这一领域的研究现状和现存问题，并提出对未来的展望。
[Key word]
[Abstract]
With the development and prevalence of artificial intelligence(AI), the application of AI has become a global trend in medical field recent years. The combination of AI and clinical medicine is becoming a hot spot in medical research. It is remarkable that the AI based on deep learning(DL)has achieved great accomplishments in image recognition, speech recognition and natural language processing, especially the application of convolutional neural network(CNN)in image recognition. CNN has been applied to the detection and diagnosis of various of ophthalmological diseases such as diabetic retinopathy, retinopathy of prematurity, glaucoma and age-related macular degeneration. Meanwhile, there are many public databases on ocular including large amount of images of fundus photographs, optical coherence tomography and visual field for researchers to train and test the AI algorithms. However, training the algorithm based on AI technology is still costly and time-consuming. The medical ethic of AI is another challengefor its application. Although DL has bought revolutionary change in ophthalmology field, it still has a long way to go for its world wide application. The article aims to synthesize the researches on the applications of AI technology in ophthalmology, and give a review on the present situation, existing problems and future outlook.
[中图分类号]
[基金项目]
深圳市第二人民医院临床研究项目(No.20193357009)
