人工智能在青光眼图像诊断中的应用  

Application of artificial intelligence in the diagnostic imaging of glaucoma

作　　者：马婧一(综述) 李元媛[2] 原慧萍(审校)[2] Ma Jingyi;Li Yuanyuan;Yuan Huiping(Harbin Medical University,Harbin 150086,China;Department of Ophthalmology,the Second Affiliated Hospital of Harbin Medical University,Harbin 150086,China)

机构地区：[1]哈尔滨医科大学,150086 [2]哈尔滨医科大学附属第二医院眼科,150086

出　　处：《中华实验眼科杂志》2020年第5期438-441,共4页Chinese Journal Of Experimental Ophthalmology

基　　金：国家重点研发计划“智能机器人”重点专项项目(2017YFB1302700);黑龙江省博士后资助项目(LBH-Z17112)。

摘　　要：随着计算机技术及图像识别技术的不断发展,人工智能(AI)在医学领域的应用不断拓展,对于利用机器学习及深度学习诊疗眼科疾病的研究日益增多。青光眼作为首位不可逆性致盲眼病,早期诊断、及时治疗对于维持患者视功能和生活质量至关重要。目前,AI在青光眼诊疗过程中的应用主要集中在青光眼辅助诊断图像,如眼底照相、视野、光相干断层扫描(OCT)等的结合,研究者们通过建立AI模型,对图像检查结果进行分割、分类及预测,达到辅助诊断及预测青光眼进展的目的。随着算法、技术的不断发展,诊断的准确性、灵敏度及特异性都在日趋完善。青光眼的诊治需要综合考量多方面辅助检查结果,因此,筛选高质量的数据并开发更加系统、全面的AI模型仍待探索。目前与青光眼相关的眼前节辅助诊断AI模型屈指可数,因此我们可以着手研究与超声生物显微镜和眼前节OCT相关的机器、深度学习模型。本文就AI在青光眼图像诊断中应用的研究进展进行综述。Along with the continuous renewal of computer technology,artificial intelligence(AI)has been widely and gradually applied in the medical field of medicine.Research concerning machine learning and deep learning for making ophthalmological diagnosis is expanding.As glaucoma is an irreversible,blinding disease,early diagnosis and treatment are extremely important for improving patients'prognosis.Currently,AI is mainly combined with several auxiliary examinations(fundus photography,visual field tests,optical coherence tomography,etc.)to diagnose and treat glaucoma.AI models are built to segment,classify,and predict the results of image examinations,which is helpful for making diagnoses and predicting glaucoma progression.With the development of algorithms and technologies,the accuracy,sensitivity,and specificity of diagnosis gradually improve.Diagnosing and treating glaucoma require comprehensive consideration of multiple auxiliary examination results,so screening high-quality data and developing more systematic and comprehensive AI models remain to be explored.Currently,only several AI models are associated with glaucoma in the anterior examinations,so it is possible to begin studying the machine and deep learning models associated with ultrasonic biological microscopy and anterior segment optical coherence tomography.This paper,therefore,reviews the application of AI in glaucoma diagnosis.

关 键 词：青光眼 人工智能 图像诊断 

分 类 号：R77[医药卫生—眼科] TP18[医药卫生—临床医学] TP391[自动化与计算机技术—控制理论与控制工程]