摘要
早产儿视网膜病变(retinopathy of prematurity,ROP)是导致儿童致盲的主要原因之一,早期ROP筛查和诊断高度依赖眼科专科医生,而随着现代医学影像技术的飞速发展和远程医疗的兴起,人工智能(artificial intelligence,AI)在ROP领域也得到了进一步应用。近年来,基于卷积神经网络、深度学习的AI在眼科筛查检测、诊疗领域得到更广泛深入地应用,其在ROP的临床诊疗方面尤引人注目,有望提高基层儿科和眼科医生对ROP早期诊断、规范治疗等方面的能力,同时降低主管医生之间的主观差异。论文概述AI在ROP自动筛查检测、诊断及预测的研究现状、存在的不足及面临的挑战,进一步了解其在ROP临床应用新动态和新进展。
Retinopathy of prematurity(ROP)is one of the main causes of blindness in children,while early screening and diagnosis of ROP are highly dependent on ophthalmologists.With the rapid development of modern medical imaging technology and the rise of telemedicine,artificial intelligence(AI)has also been further applied in the field of ROP.In recent years,AI based on neural convolutional network and deep learning has been applied in the field of eye screening,disease detection,diagnosis and treatment,especially for the clinical diagnosis and treatment of ROP.It is expected that AI will aid primary pediatricians and ophthalmologists in early diagnosis and standardized treatment of ROP,and reduce the subjective differences among first-line doctors.This review summarizes the current status,shortcomings,and challenges faced by AI in automated ROP screening,diagnosis,and prediction,which enables the further understanding of new trends and progressment made by its clinical application in ROP.
作者
杜曰山一
王鲜
张国明
DU Yueshanyi;WANG Xian;ZHANG Guoming(Clinical Medical College,Guizhou Medical University,Guiyang 550004,Guizhou,China;Shenzhen Eye Hospital,Jinan University,Shenzhen Eye Institute,Shenzhen 518040,Guangdong,China;Department of Ophthalmology,Affiliated Hospital of Guizhou Medical University,Guiyang 550004,Guizhou,China)
出处
《山东大学耳鼻喉眼学报》
CAS
2023年第3期157-162,共6页
Journal of Otolaryngology and Ophthalmology of Shandong University
基金
国家自然科学基金资助项目(82271103)
广东省基础与应用基础研究基金项目(2022A1515012326)
深圳市医学重点学科建设经费资助项目(SZXK038)
广东省高水平临床重点专科(深圳市配套建设经费)资助项目(SZGSP014)
深港联合资助项目(A类)(SGDX20190920110403741)。
关键词
早产儿视网膜病变
人工智能
深度学习
自动筛查
诊断
预测
Retinopathy of prematurity
Artificial intelligence
Deep learning
Automated screening
Diagnosis
Prediction
