EN / 中文
您的位置： 首页 > 2022年10月 第37卷 第10期 > 文字全文
2023年7月 第38卷 第7期11
目录

人工智能与婴幼儿眼病的研究进展

Advances in application of artificial intelligence in infantile eye diseases

来源期刊： 眼科学报 | 2022年10月 第37卷 第10期： 799-805 发布时间： 收稿时间：2022/11/21 8:09:34 阅读量：3968
作者：
杨雪莹,
李建桥,
关键词：
人工智能深度学习卷积神经网络婴幼儿眼病
artificial intelligence deep learning convolution neural network infant eye diseases
DOI：
10.3978/j.issn.1000-4432.2022.04.08
收稿时间：
 
修订日期：
 
接收日期：
 
近年来人工智能(artificial intelligence，AI)技术在医学领域的应用发展迅猛，尤其在眼科领域，成果显著，极大地提高了相关影像数据的诊断效率，推动了该领域研究的进展。然而，大多数AI的应用都集中于成人眼病，在婴幼儿眼病方向的研究较少。究其原因，可能是婴幼儿眼部影像数据采集配合度低，部分影像设备应用受限，且相关领域专业眼科医生数量匮乏。然而，婴幼儿期是视觉发育最重要的阶段，也是出生缺陷早期筛防诊治的重灾区，对患儿的视觉发展具有长远且重要的影响，亟需AI相关产品提高婴幼儿眼病筛查效率，缓解医疗资源不足的现状。本文将对近年AI在婴幼儿眼病领域的研究应用现状、进展及存在的相关问题进行综述。
In recent years, the application of artificial intelligence (AI) in medicine, especially in ophthalmology, has developed rapidly with remarkable results. This has greatly improved the diagnostic efficiency of relevant imaging data and promoted further research in this field. However, most applications of AI are focused on adult eye diseases, and few studies have addressed infantile eye diseases. This may be because of the non-cooperative nature of infants, the limited availability of imaging equipment in infants, and the lack of pediatric ophthalmologists. Infancy is the most important stage of vision development. Disturbance during this period have a profound and lasting influence on vision development. Hence, early screening, diagnosis, and treatment of birth defects is important. AI-related products, which improves the efficacy of infant eye disease screening, are urgently needed. This paper reviews the current status, progress, and existing problems of recent research related to application of AI in infantile eye diseases.
婴幼儿时期是眼部疾病的高发期[1]，早产儿视网膜病变(retinopathy of prematurity，ROP)、先天性白内障(congenital cataract，CC)及视网膜母 细胞瘤(retinoblastoma，RB)等疾病可导致婴幼儿发生不可逆性视力损伤。由于患儿不具备语言功能，且眼部症状不明显，常常错过最佳诊疗期， 因此，早期筛查与诊断至关重要。但现有医疗资源分布不均、专业儿童眼科医生缺乏[2]，使得大范围筛查工作面临挑战。此外，处于快速发育期的婴幼儿眼球结构变化较大，往往需要多次跟踪随访，而传统的人工筛查方式费时费力，且主观性强，极易漏诊或误诊，加重视力损害负担。基于数据分析的人工智能(artificial intelligence，AI) 技术与眼科影像学的结合，可能为解决上述问题 提供新思路。AI通过对临床影像资料进行分析处理，从大数据中学习经验，进行自动化图像识别与疾病诊断，在促进客观、便捷、高效、全面的婴幼儿眼病早期筛查方面具有重要意义。因此本文旨在对AI与婴幼儿眼病领域的研究应用现状、进展及存在的相关问题进行分析总结，并对其未来发展做出展望。

1 AI 概述

      AI的概念最早于1956年提出，是利用电子计算机模拟人类智力活动的技术[3]。机器学习 (machine learning，ML)是指从数据中自动学习的 A I，在医疗保健领域，输入通常是医学图像，通 过映射函数输出对特定疾病的诊断[4]，在大数据分析中得到广泛应用；深度学习(deep learning， DL)是ML子领域[5]，用于分类与特征提取，涉及具有多层处理的训练模型。卷积神经网络 (convolution neural network，CNN)是一种常用的 进行图像数据识别的深度网络，通过重复和校正过程进行自我学习，分析专家标注的图像标签训练集进行诊断[4]，在检测糖尿病性视网膜病变[6]、 青光眼[7]、年龄相关性黄斑变性[8]以及通过数字眼底照片识别心血管危险因素和疾病方面已被验证是准确有效的[9]。AI作为医生诊疗过程中的辅 助工具，在客观诊断、远程医疗、优化患者护理、降低成本等方面发挥重要作用。

2 AI

      在婴幼儿眼病中的应用 2.1 AI 与 ROP ROP是一种视网膜发育异常引起的血管增殖性疾病[10]，是全球范围内儿童失明的主要原因[11]。 早期诊断和及时治疗是控制疾病进展的关键，然而，传统使用间接检眼镜[12]的人工筛查方法耗时耗力，且主观性强[13]，在疾病的大范围筛查和 标准化诊断过程中面临挑战。随着科技的发展， 新生儿数字化广域眼底成像系统的出现为疾病 的客观诊断带来福音，人们开始尝试将其与DL 结合建立自动ROP筛查工具。Huang等[14]提出了1个检测早期ROP的DL模型，其在测试集中可实现92.23%的准确率，能够有效避免漏诊和治疗时机的延误。Wang等[15]建立了2种特定的深度神 经网络Id-Net和Gr-Net，分别用于对ROP诊断和 轻重度分级，并将此算法集成到云平台DeepROP 中，实现自动ROP检测和报告生成、样本下载及 报告评估，在实际应用中，模型的诊断效能甚 至超越了一些人类专家。Peng等[16]提出了基于 ResNet18、DenseNet121和EfficientNetB2三种并行框架的自动ROP5级分期网络，通过拼接和卷积将 特征提取深度融合，并采取有序分类策略，提高 了模型的性能。
      除了早期筛查，具有“后极部视网膜静脉的 扩张和动脉的弯曲”特征的附加病变作为ROP严重程度的标志，一直是研究的热点。早前，AtaerCansizoglu等[17]利用主成分生成森林( principal spanning forest)算法开发了广域视网膜图像分析系 统“i-ROP”，该系统从视网膜动脉和静脉中提取弯曲和扩张特征以诊断附加病变，准确率可高达 95%；在此基础上，Brown等[18]建立了CNN系统来 预测ROP的三级分类(正常、附加前病变、附加病变)，该系统包括1个具有U-net结构的眼底图像血管分割处理网络和1个具有Inception V1结构的分类网络，对附加病变的诊断可达到93%的敏感性和 94%的特异性，甚至在测试集中表现超过了6/8的 专家。此外，I-ROP ASSIST[19]、RO P. AI[20]等模 型也被提出用于附加病变的诊断，作为一种筛查工具，这些AI模型的表现性能预示着其在附加病 变诊断方面的应用价值。随后，Tong等[21]开发了包含ResNet和Faster-RCNN两个神经网络的DL系统，基于图像特征进行ROP严重程度的自动分级，并实现了对疾病分期和附加病变的进一步客 观高效诊断。 
      然而，有学者[22]在后续工作中发现，仅关注附加病变不足以从整体上反映疾病的严重程度，进而提出了一种基于i-ROP系统的视网膜血管严重 程度自动评分系统，纳入了分区、分期、附加病 变这3个关键的诊断参数，自动区分无ROP，轻度 ROP，2型ROP以及1型ROP，具有极高的敏感性和特异性。随着医学不断地向客观和定量诊断的方向发展，这种DL衍生的ROP评分系统在后续护理 工作中显示了强大的应用价值，不但可以跟踪监 测疾病进展[23]，准确识别出需要转诊的病例，对急进型ROP(aggressive ROP，A-ROP)[24]这种快速 进展的特殊类型也做出了定量描述，甚至在激光 或抗血管内皮生长因子治疗后疾病消退及复发情 况的追踪方面也表现出了强大的效能[25]。最近， Campbell等[13]在先前工作的基础上对血管严重程 度进行了更细致的1~9评分，帮助临床专家识别出与疾病严重程度相关的血管细微的改变，辅助 筛查的同时提高了人类医生对疾病进展的认识。 Wang等[26]进一步开发了1个J-PROP深度学习平台，整合了图像质量、ROP分期、眼内出血、附加前病变/附加病变多个维度，综合眼底图像的多 维病理病变进行诊断并做出转诊建议，促进了更加精确全面的ROP筛查诊断过程。

2.2 AI 与 CC

      CC指在出生或儿童早期发生的晶状体混浊， 是世界范围内可避免的儿童失明的主要原因之 一[27]。在视觉系统发育的敏感时期，CC往往导致不可逆的视觉缺陷，早期诊断和治疗对防止形觉剥夺型弱视的发展至关重要[28]。因此，近年研究开始将AI引入白内障的诊断过程，构建自动CC筛查模型，提高诊疗效率。Long等[29]通过对眼前段裂隙灯照片进行训练，建立了AI模型CCCruiser，包括3个独立的CNN功能网络：用以从海量人群中诊断出CC的识别网络，根据晶体混浊区域、密度、位置对疾病严重程度进行分层的评估网络，以及提供参考治疗方案的决策网络。并建立云平台实现数据的整合与共享，促进了筛查的高效进行。多中心随机对照试验[30]显示：虽然在实际工作中模型的准确率低于专家，但患者对 其快速评估的满意度很高。然而，上述A I模型是使用专业眼科设备收集的图像数据进行训练的，这意味着在医疗资源不足的地区，其用途可能有限，因此，Lin等[27]采用随机森林(random forest RF)和自适应增强(adaptive boosting)算法，基于家 族史、家庭环境、新生儿合并症等11个不基于图 像的易获取的风险因素训练了1个CC识别模型，用以区分CC患者和健康儿童，在临床试验中表现出较强的鲁棒性，作为一种辅助筛查工具，特别是在偏远地区，该模型有潜力弥补目前CC筛查的不足，对早期检测或预测CC的发展具有重要意义。 
      由于婴幼儿眼球结构处于快速发育的时期，术中人工晶体(intraocular lens，IOL)度数的选择成 为一大难题[31]。已有研究[32]表明ML算法有助于提高成人IOL计算公式的性能并优化晶体的选择，但其在婴幼儿IOL测算领域有待进一步研究。除度数要求较高外，晶体植入的时机也是影响视力预后的关键因素[33]。即使在术后，人工晶体眼的儿童也因其高并发症发生率而需要长期监测以防止弱视的发生[34]。Zhang等[35]将RF和朴素贝叶斯(Na?ve Bayesian，NB)算法应用于人口统计学信息和白内障评估数据，建立了高眼压和中央晶状体再生两种 术后常见并发症的风险预测模型，对威胁视力的并发症进行预警，平均准确率可达75%。Long等[36] 采用贝叶斯(Bayesian)算法创建了CC-Guardian用于 CC术后的随访管理，根据12个输入变量预测上述2种并发症，该模型集成了预测、分配调度、远程 医疗3个模块，能够准确识别高危患儿并做出后续额外治疗的临床决定。且该模型与智能手机App和远程预测云平台的结合，与传统方法相比，节约了医疗成本，具有显著的社会经济效益。

2.3 AI 与斜弱视

      弱视是视觉发育期内由于异常视觉经验引起 的单眼或双眼最佳矫正视力下降，且眼部检查无器质性病变，发生在1%~4%的儿童中[37]，是婴幼儿眼科筛查项目中最重要的工作之一。对于3岁以下的婴幼儿，视力筛查通常采用固视和追随实验[38]，然而该过程耗时耗力，具有很强的局限性。为了克服综合视觉评估的困难，Pueyo等[38]通 过集成视觉检查设备中经过训练的AI网络，结合眼动追踪技术收集到的凝视数据，对6个月及以上 的婴幼儿进行快速视力筛查，以降低视力损害的 发生率。 
      斜视是发展为弱视的风险因素之一，儿童患病率约为2%~4%[39]。斜视不仅干扰双眼视功能，还将对患儿产生持久的社会心理影响[40]。此外，斜视也可能是白内障、青光眼，以及视神经、眼眶或大脑肿瘤的第一症状。因此早期诊断对挽救视功能至关重要，但常规的遮盖实验在3岁以下婴 幼儿中准确性较差。Zheng等[41]利用Faster R-CNN 对眼部凝视图片进行自动裁剪，训练深度卷积神经网络(deep convolution neural network，DCNN)模型使其能够自动识别水平斜视，准确性可达95%，并优于专家水平，为斜视患儿的诊断及转诊提供 了便捷。

2.4 AI 与 RB

      RB是最常见的眼内恶性肿瘤，在5岁以下婴幼儿中发病率最高[42]，可达14.1/100万，世界范围内每年估计有9000名新确诊患者，由于恶性肿瘤生长和转移迅速，大多数病例常因为救治无效而死亡[43]。因此，早期诊断和及时治疗是最大限度提高生存机会和挽救视力的关键[44]。临床工作中， 通常使用磁共振成像(magnetic resonance imaging，MRI)来评估病变范围及神经转移情况[45]，但进行个体化眼结构与肿瘤组织分割的人工圈定过程较为繁琐，且难以量化。因此，有学者提出将MRI与 AI相结合进行自动化分割，Ciller等[46]利用主动形 状模型(Active Shape Models，ASM)结合3DCNN对病理眼的MRI数据进行训练，使其能够自动描绘巩膜、角膜、晶状体、玻璃体和肿瘤的区域，在病理组织分割方面显示出较强的鲁棒性和有效性。随后，Strijbis等[47]建立多视图卷积神经网络(multiview convolutional neural networks，MV-CNNs)旨在区分正常眼内结构和肿瘤，并对瘤体体积进行定量描述，与传统方法相比，MV-CNNs可实现更高的分割精度，显示出优越的体积和空间分割性能，具有很大的发展潜力。

2.5 AI 与婴幼儿眼病常见的体征

2.5.1 AI 与眼底出血

      眼底出血是新生儿最常见的眼部异常体征之一，一项眼筛研究[48]显示其占比可达88.9%。与 出生有关的出血往往体积小、数量少且很快自发消失[49]，而那些出血范围大、持续时间长或位 置特殊的情况可能是脑损伤的表现[50]，延误治疗会引起弱视及出血相关的视网膜毒性损害，影 响患儿视力发育。Wang等[49]在对新生儿眼底图片训练的基础上建立了1个DCNN模型进行眼底出血的诊断和1~3级严重程度分级，准确率可达 97.85%~99.96%。为进一步对眼底出血进行更加客观、综合的描述，Mao等[51]基于DCNN对出血区域的高精度分割结果，定义了一种包含出血面积 与视盘比例以及出血相对于黄斑区位置的新生儿出血分级标准，旨在对疾病严重程度进行全面评估，为今后临床决策提供定量参考依据，指导医生进行临床诊断。

2.5.2 AI 与白瞳症

      白瞳症是一种不正常的瞳孔反射，在缺乏语言表达能力的患儿身上，其常常是RB、白内障、ROP、永存原始玻璃体增生症、眼弓蛔虫感染[52] 等婴幼儿眼病的首发表现。传统的新生儿视力筛查工作由初级眼保健医生使用红光反射进行，然而，散瞳不完全、婴幼儿依从性差等原因使得这 种方式对于异常体征的检出能力有限。有学者[53] 提出：由反向传播算法(backpropagation，BP)训 练构建的CNN可以在智能手机上自主运行，应用程序CRADLE(ComputeR-Assisted Detector of Leukocoria)通过分析储存在用户移动设备中的婴幼儿图片辅助检测白瞳症。相比于红光反射，拍 摄照片更能作为一种频繁的筛查方式提高呈现低亮度和低分辨率的病理性白瞳的检出率，减少漏诊。在后续测试中，CRADLE成功在临床诊断前1.3年检测到了白瞳症，作为婴幼儿眼病筛查的补 充工具，CRADLE在提示患儿及时转诊、改善视力 预后方面显示出了重要意义。

2.6 AI 与婴幼儿的其他视觉方面

      早前，Zhang等[54]利用反向传播神经网络预测了IOL植入术后各年龄的屈光状态，可用于指导婴幼儿白内障手术。早期进行白内障手术的患儿面 部识别能力有所下降，会出现类似“脸盲”的表现。为了解释这一现象，Vogelsang等[55]在大型脸部图像数据库训练的基础上建立了1个CNN系统， 来观察不同视敏度水平的发展对空间整合和人脸识别功能的影响，结果显示：接受CC手术IOL植入的患儿在早期获得了高初始视敏度，打破了低初始视敏度诱导构型面部识别皮层空间加工过程的生理发育规律，反而在轮廓知觉和整体形状完 成方面受到损害。而在这之前，人们一直将这种现象归咎于早期白内障造成的形觉剥夺引起的面部处理网络连接活动的中断[56]。以上应用进一步表明，AI在辅助诊疗的同时也促进了人类医生对疾病的认识。

3 结语

      随着AI日渐融入婴幼儿眼病领域，从疾病诊断到术后随访，推动了传统医疗方式的变革，不仅以较高的诊断准确率为更大范围客观高效筛查提供机会，缓解了医疗机构的压力，且有望通过云平台进行数据整合解决世界范围内医疗资源分布不均的问题。此外，多元化电子产品终端服务 器如智能手机的普及，亦将有效整合A I、远程医疗等相关线上平台的服务能力，促进筛查效率， 完善疾病的二级预防。尤其是在新冠疫情期间传 统筛查随访模式受到巨大冲击的情况下，基于智能手机的远程医疗平台的应用[57-58]，凸显了其在辅 助婴幼儿眼病筛查方面的重要价值。同时该技术也为年轻医生提供了提高诊疗技能的学习平台， 具有可观的发展前景和社会效益。
      但在AI给医疗领域带来便捷的同时也应考虑，此项技术在婴幼儿眼病的应用尚不成熟：1)AI 算法的基础相对机械，相关模型也未能做到模态融 合，纳入更多评估因素进行多模态分析将有助于开 展大范围的筛查工作；2)眼底图像拍摄的差异、缺 乏标准化的图像采集流程等问题使得模型的诊断准 确率受限，应加快婴幼儿眼病标准化数据库建设，促进资源共享，提高ML模型的性能；3)目前的AI 系统大多为单病种诊断，缺乏整合眼前节或后节的 多病种筛查模型，易漏诊Coats病、有髓鞘视神经纤维等在智能筛查领域仍处于空白的眼病，因此， 应扩大研究的病种领域，建立更加系统的筛查预测模型，促进综合的疾病诊断过程。
      AI在婴幼儿眼病领域的应用潜力巨大，随着技术的创新升级和互联网时代的发展，AI有望以更高的诊断准确率、更广的病种覆盖率，促进筛查过程中的自动化与智能化，在目前研究成果的基础上实现新的突破，为更多患儿带来希望。

开放获取声明

    本文适用于知识共享许可协议 (Creative Commons)，允许第三方用户按照署名(BY)-非商业性使用(NC)-禁止演绎(ND)(CC BY-NC-ND)的方式共享，即允许第三方对本刊发表的文章进行复制、发行、展览、表演、放映、广播或通过信息网络向公众传播，但在这些过程中必须保留作者署名、仅限于非商业性目的、不得进行演绎创作。详情请访问：https://creativecommons.org/licenses/by-nc-nd/4.0/。
1、Teoh LJ, Solebo AL, Rahi JS, et al. Temporal trends in the epidemiology of childhood severe visual impairment and blindness in the UK[ J/OL]. Br J Ophthalmol, 2021, Epub ahead of print.Teoh LJ, Solebo AL, Rahi JS, et al. Temporal trends in the epidemiology of childhood severe visual impairment and blindness in the UK[ J/OL]. Br J Ophthalmol, 2021, Epub ahead of print.
2、Reid JE, Eaton E. Artificial intelligence for pediatric ophthalmology[ J]. Curr Opin Ophthalmol, 2019, 30(5): 337-346.Reid JE, Eaton E. Artificial intelligence for pediatric ophthalmology[ J]. Curr Opin Ophthalmol, 2019, 30(5): 337-346.
3、Li JO, Liu H, Ting DSJ, et al. Digital technology, tele-medicine and artificial intelligence in ophthalmology: A global perspective[ J]. Prog Retin Eye Res, 2021, 82: 100900.Li JO, Liu H, Ting DSJ, et al. Digital technology, tele-medicine and artificial intelligence in ophthalmology: A global perspective[ J]. Prog Retin Eye Res, 2021, 82: 100900.
4、Dutt S, Sivaraman A, Savoy F, et al. Insights into the growing popularity of artificial intelligence in ophthalmology[ J]. Indian J Ophthalmol, 2020, 68(7): 1339-1346.Dutt S, Sivaraman A, Savoy F, et al. Insights into the growing popularity of artificial intelligence in ophthalmology[ J]. Indian J Ophthalmol, 2020, 68(7): 1339-1346.
5、Hanif AM, Beqiri S, Keane PA, et al. Applications of interpretability in deep learning models for ophthalmology[ J]. Curr Opin Ophthalmol, 2021, 32(5): 452-458.Hanif AM, Beqiri S, Keane PA, et al. Applications of interpretability in deep learning models for ophthalmology[ J]. Curr Opin Ophthalmol, 2021, 32(5): 452-458.
6、Giger ML. Machine learning in medical imaging[ J]. J Am Coll Radiol, 2018, 15(3 Pt B): 512-520.Giger ML. Machine learning in medical imaging[ J]. J Am Coll Radiol, 2018, 15(3 Pt B): 512-520.
7、Liu H, Li L, Wormstone IM, et al. Development and validation of a deep learning system to detect glaucomatous optic neuropathy using fundus photographs[ J]. JAMA Ophthalmol, 2019, 137(12): 1353-1360.Liu H, Li L, Wormstone IM, et al. Development and validation of a deep learning system to detect glaucomatous optic neuropathy using fundus photographs[ J]. JAMA Ophthalmol, 2019, 137(12): 1353-1360.
8、Grassmann F, Mengelkamp J, Brandl C, et al. A deep learning algorithm for prediction of age-related eye disease study severity scale for age related macular degeneration from color fundus photography[ J]. Ophthalmology, 2018, 125(9): 1410-1420.Grassmann F, Mengelkamp J, Brandl C, et al. A deep learning algorithm for prediction of age-related eye disease study severity scale for age related macular degeneration from color fundus photography[ J]. Ophthalmology, 2018, 125(9): 1410-1420.
9、Ting DSW, Peng L, Varadarajan AV, et al. Deep learning in ophthalmology: the technical and clinical considerations[ J]. Prog Retin Eye Res, 2019, 72: 100759.Ting DSW, Peng L, Varadarajan AV, et al. Deep learning in ophthalmology: the technical and clinical considerations[ J]. Prog Retin Eye Res, 2019, 72: 100759.
10、Sen P, Wu WC, Chandra P, et al. Retinopathy of prematurity treatment: Asian perspectives[ J]. Eye (Lond), 2020, 34(4): 632-642.Sen P, Wu WC, Chandra P, et al. Retinopathy of prematurity treatment: Asian perspectives[ J]. Eye (Lond), 2020, 34(4): 632-642.
11、Wu T, Zhang L, Tong Y, et al. Retinopathy of prematurity among very low-birth-weight infants in China: incidence and perinatal risk factors[ J]. Invest Ophthalmol Vis Sci, 2018, 59(2): 757-763.Wu T, Zhang L, Tong Y, et al. Retinopathy of prematurity among very low-birth-weight infants in China: incidence and perinatal risk factors[ J]. Invest Ophthalmol Vis Sci, 2018, 59(2): 757-763.
12、Vartanian RJ, Besirli CG, Barks JD, et al. Trends in the screening and treatment of retinopathy of prematurity[ J]. Pediatrics, 2017, 139(1): e20161978.Vartanian RJ, Besirli CG, Barks JD, et al. Trends in the screening and treatment of retinopathy of prematurity[ J]. Pediatrics, 2017, 139(1): e20161978.
13、Campbell JP, Kim SJ, Brown JM, et al. Evaluation of a deep learning-derived quantitative retinopathy of prematurity severity scale[ J]. Ophthalmology, 2021, 128(7): 1070-1076.Campbell JP, Kim SJ, Brown JM, et al. Evaluation of a deep learning-derived quantitative retinopathy of prematurity severity scale[ J]. Ophthalmology, 2021, 128(7): 1070-1076.
14、Huang YP, Basanta H, Kang EY, et al. Automated detection of early-stage ROP using a deep convolutional neural network[ J]. Br J Ophthalmol, 2021, 105(8): 1099-1103.Huang YP, Basanta H, Kang EY, et al. Automated detection of early-stage ROP using a deep convolutional neural network[ J]. Br J Ophthalmol, 2021, 105(8): 1099-1103.
15、Wang J, Ju R, Chen Y, et al. Automated retinopathy of prematurity screening using deep neural networks[ J]. EBioMedicine, 2018, 35: 361-368.Wang J, Ju R, Chen Y, et al. Automated retinopathy of prematurity screening using deep neural networks[ J]. EBioMedicine, 2018, 35: 361-368.
16、Peng Y, Zhu W, Chen Z, et al. Automatic staging for retinopathy of prematurity with deep feature fusion and ordinal classification strategy[ J]. IEEE Trans Med Imaging, 2021, 40(7): 1750-1762.Peng Y, Zhu W, Chen Z, et al. Automatic staging for retinopathy of prematurity with deep feature fusion and ordinal classification strategy[ J]. IEEE Trans Med Imaging, 2021, 40(7): 1750-1762.
17、Ataer-Cansizoglu E, Bolon-Canedo V, Campbell JP, et al. Computer-based image analysis for plus disease diagnosis in retinopathy of prematurity: performance of the "i-ROP" system and image features associated with expert diagnosis[ J]. Transl Vis Sci Technol, 2015, 4(6): 5.Ataer-Cansizoglu E, Bolon-Canedo V, Campbell JP, et al. Computer-based image analysis for plus disease diagnosis in retinopathy of prematurity: performance of the "i-ROP" system and image features associated with expert diagnosis[ J]. Transl Vis Sci Technol, 2015, 4(6): 5.
18、Brown JM, Campbell JP, Beers A, et al. Automated diagnosis of plus disease in retinopathy of prematurity using deep convolutional neural networks[ J]. JAMA Ophthalmol, 2018, 136(7): 803-810.Brown JM, Campbell JP, Beers A, et al. Automated diagnosis of plus disease in retinopathy of prematurity using deep convolutional neural networks[ J]. JAMA Ophthalmol, 2018, 136(7): 803-810.
19、Yildiz VM, Tian P, Yildiz I, et al. Plus disease in retinopathy of prematurity: convolutional neural network performance using a combined neural network and feature extraction approach[ J]. Transl Vis Sci Technol, 2020, 9(2): 10.Yildiz VM, Tian P, Yildiz I, et al. Plus disease in retinopathy of prematurity: convolutional neural network performance using a combined neural network and feature extraction approach[ J]. Transl Vis Sci Technol, 2020, 9(2): 10.
20、Tan Z, Simkin S, Lai C, et al. Deep learning algorithm for automated diagnosis of retinopathy of prematurity plus disease[ J]. Transl Vis Sci Technol, 2019, 8(6): 23.Tan Z, Simkin S, Lai C, et al. Deep learning algorithm for automated diagnosis of retinopathy of prematurity plus disease[ J]. Transl Vis Sci Technol, 2019, 8(6): 23.
21、Tong Y, Lu W, Deng QQ, et al. Automated identification of retinopathy of prematurity by image-based deep learning[ J]. Eye Vis (Lond), 2020, 7: 40.Tong Y, Lu W, Deng QQ, et al. Automated identification of retinopathy of prematurity by image-based deep learning[ J]. Eye Vis (Lond), 2020, 7: 40.
22、Redd TK , Campbell JP, Brown JM, et al. Evaluation of a deep learning image assessment system for detecting severe retinopathy of prematurity[ J/OL]. Br J Ophthalmol, 2018, Epub ahead of print.Redd TK , Campbell JP, Brown JM, et al. Evaluation of a deep learning image assessment system for detecting severe retinopathy of prematurity[ J/OL]. Br J Ophthalmol, 2018, Epub ahead of print.
23、Taylor S, Brown JM, Gupta K, et al. Monitoring disease progression with a quantitative severity scale for retinopathy of prematurity using deep learning[ J]. JAMA Ophthalmol, 2019, 137(9): 1022-1028.Taylor S, Brown JM, Gupta K, et al. Monitoring disease progression with a quantitative severity scale for retinopathy of prematurity using deep learning[ J]. JAMA Ophthalmol, 2019, 137(9): 1022-1028.
24、Bellsmith KN, Brown J, Kim SJ, et al. Aggressive posterior retinopathy of prematurity: clinical and quantitative imaging features in a large north American cohort[ J]. Ophthalmology, 2020, 127(8): 1105-1112.Bellsmith KN, Brown J, Kim SJ, et al. Aggressive posterior retinopathy of prematurity: clinical and quantitative imaging features in a large north American cohort[ J]. Ophthalmology, 2020, 127(8): 1105-1112.
25、Gupta K, Campbell JP, Taylor S, et al. A quantitative severity scale for retinopathy of prematurity using deep learning to monitor disease regression after treatment[ J/OL]. JAMA Ophthalmol, 2019, Epub ahead of print.Gupta K, Campbell JP, Taylor S, et al. A quantitative severity scale for retinopathy of prematurity using deep learning to monitor disease regression after treatment[ J/OL]. JAMA Ophthalmol, 2019, Epub ahead of print.
26、Wang J, Ji J, Zhang M, et al. Automated explainable multidimensional deep learning platform of retinal images for retinopathy of prematurity screening[ J]. JAMA Netw Open, 2021, 4(5): e218758.Wang J, Ji J, Zhang M, et al. Automated explainable multidimensional deep learning platform of retinal images for retinopathy of prematurity screening[ J]. JAMA Netw Open, 2021, 4(5): e218758.
27、Lin D, Chen J, Lin Z, et al. A practical model for the identification of congenital cataracts using machine learning[ J]. EBioMedicine, 2020, 51: 102621.Lin D, Chen J, Lin Z, et al. A practical model for the identification of congenital cataracts using machine learning[ J]. EBioMedicine, 2020, 51: 102621.
28、Bremond-Gignac D, Daruich A, Robert MP, et al. Recent developments in the management of congenital cataract[ J]. Ann Transl Med, 2020, 8(22): 1545.Bremond-Gignac D, Daruich A, Robert MP, et al. Recent developments in the management of congenital cataract[ J]. Ann Transl Med, 2020, 8(22): 1545.
29、Long E, Lin H, Liu Z, et al. An artificial intelligence platform for the multihospital collaborative management of congenital cataracts[ J]. Nat Biomed Eng, 2017, 24: 1.Long E, Lin H, Liu Z, et al. An artificial intelligence platform for the multihospital collaborative management of congenital cataracts[ J]. Nat Biomed Eng, 2017, 24: 1.
30、Lin H, Li R, Liu Z, et al. Diagnostic efficacy and therapeutic decision making capacity of an artificial intelligence platform for childhood cataracts in eye clinics: a multicentre randomized controlled trial[ J]. EClinicalMedicine, 2019, 9: 52-59.Lin H, Li R, Liu Z, et al. Diagnostic efficacy and therapeutic decision making capacity of an artificial intelligence platform for childhood cataracts in eye clinics: a multicentre randomized controlled trial[ J]. EClinicalMedicine, 2019, 9: 52-59.
31、Infant Aphakia Treatment Study Group; Lambert SR, Lynn MJ, et al. Comparison of contact lens and intraocular lens correction of monocular aphakia during infancy: a randomized clinical trial of HOTV optotype acuity at age 4.5 years and clinical findings at age 5 years[ J]. JAMA Ophthalmol, 2014, 132(6): 676-682.Infant Aphakia Treatment Study Group; Lambert SR, Lynn MJ, et al. Comparison of contact lens and intraocular lens correction of monocular aphakia during infancy: a randomized clinical trial of HOTV optotype acuity at age 4.5 years and clinical findings at age 5 years[ J]. JAMA Ophthalmol, 2014, 132(6): 676-682.
32、Sramka M, Slovak M, Tuckova J, et al. Improving clinical refractive results of cataract surgery by machine learning[ J]. PeerJ, 2019, 7: e7202.Sramka M, Slovak M, Tuckova J, et al. Improving clinical refractive results of cataract surgery by machine learning[ J]. PeerJ, 2019, 7: e7202.
33、Lin D, Liu Z, Chen J, et al. Practical pattern of surgical timing of childhood cataract in China: A cross-sectional database study[ J]. Int J Surg, 2019, 62: 56-61.Lin D, Liu Z, Chen J, et al. Practical pattern of surgical timing of childhood cataract in China: A cross-sectional database study[ J]. Int J Surg, 2019, 62: 56-61.
34、Bothun ED, Wilson ME, Traboulsi EI, et al. Outcomes of unilateral cataracts in infants and toddlers 7 to 24 months of age: toddler aphakia and pseudophakia study (TAPS)[ J]. Ophthalmology, 2019, 126(8): 1189-1195.Bothun ED, Wilson ME, Traboulsi EI, et al. Outcomes of unilateral cataracts in infants and toddlers 7 to 24 months of age: toddler aphakia and pseudophakia study (TAPS)[ J]. Ophthalmology, 2019, 126(8): 1189-1195.
35、Zhang K, Liu X, Jiang J, et al. Prediction of postoperative complications of pediatric cataract patients using data mining[ J]. J Transl Med, 2019, 17(1): 2.Zhang K, Liu X, Jiang J, et al. Prediction of postoperative complications of pediatric cataract patients using data mining[ J]. J Transl Med, 2019, 17(1): 2.
36、Long E, Chen J, Wu X, et al. Artificial intelligence manages congenital cataract with individualized prediction and telehealth computing[ J]. NPJ Digit Med, 2020, 3: 112.Long E, Chen J, Wu X, et al. Artificial intelligence manages congenital cataract with individualized prediction and telehealth computing[ J]. NPJ Digit Med, 2020, 3: 112.
37、Loh AR, Chiang MF. Pediatric vision screening[ J]. Pediatr Rev, 2018, 39(5): 225-234.Loh AR, Chiang MF. Pediatric vision screening[ J]. Pediatr Rev, 2018, 39(5): 225-234.
38、Pueyo V, Pérez-Roche T, Prieto E, et al. Development of a system based on artificial intelligence to identify visual problems in children: study protocol of the TrackAI project[ J]. BMJ Open, 2020, 10(2): e033139. Pueyo V, Pérez-Roche T, Prieto E, et al. Development of a system based on artificial intelligence to identify visual problems in children: study protocol of the TrackAI project[ J]. BMJ Open, 2020, 10(2): e033139.
39、Hatt SR, Leske DA, Casta?eda YS, et al. Association of strabismus with functional vision and eye-related quality of life in children[ J]. JAMA Ophthalmol, 2020, 138(5): 528-535.Hatt SR, Leske DA, Casta?eda YS, et al. Association of strabismus with functional vision and eye-related quality of life in children[ J]. JAMA Ophthalmol, 2020, 138(5): 528-535.
40、Hull S, Tailor V, Balduzzi S, et al. Tests for detecting strabismus in children aged 1 to 6 years in the community[ J]. Cochrane Database Syst Rev, 2017, 11: CD011221.Hull S, Tailor V, Balduzzi S, et al. Tests for detecting strabismus in children aged 1 to 6 years in the community[ J]. Cochrane Database Syst Rev, 2017, 11: CD011221.
41、Zheng C, Yao Q, Lu J, et al. Detection of referable horizontal strabismus in children's primary gaze photographs using deep learning[ J]. Transl Vis Sci Technol, 2021, 10(1): 33.Zheng C, Yao Q, Lu J, et al. Detection of referable horizontal strabismus in children's primary gaze photographs using deep learning[ J]. Transl Vis Sci Technol, 2021, 10(1): 33.
42、Stacey AW, Bowman R, Foster A, et al. Incidence of retinoblastoma has increased: results from 40 European countries[ J]. Ophthalmology, 2021, 128(9): 1369-1371.Stacey AW, Bowman R, Foster A, et al. Incidence of retinoblastoma has increased: results from 40 European countries[ J]. Ophthalmology, 2021, 128(9): 1369-1371.
43、Dimaras H, Kimani K, Dimba EA, et al. Retinoblastoma[ J]. Lancet, 2012, 379(9824): 1436-1446.Dimaras H, Kimani K, Dimba EA, et al. Retinoblastoma[ J]. Lancet, 2012, 379(9824): 1436-1446.
44、Global Retinoblastoma Study Group; Fabian ID, Abdallah E, et al. Global retinoblastoma presentation and analysis by national income level[ J]. JAMA Oncol, 2020, 6(5): 685-695.Global Retinoblastoma Study Group; Fabian ID, Abdallah E, et al. Global retinoblastoma presentation and analysis by national income level[ J]. JAMA Oncol, 2020, 6(5): 685-695.
45、Jansen RW, de Bloeme CM, Brisse HJ, et al. MR imaging features to differentiate retinoblastoma from Coats' disease and persistent fetal vasculature[ J]. Cancers (Basel), 2020, 12(12): 3592.Jansen RW, de Bloeme CM, Brisse HJ, et al. MR imaging features to differentiate retinoblastoma from Coats' disease and persistent fetal vasculature[ J]. Cancers (Basel), 2020, 12(12): 3592.
46、Ciller C, De Zanet S, Kamnitsas K , et al. Multi-channel MRI segmentation of eye structures and tumors using patient-specific features[ J]. PLoS One, 2017, 12(3): e0173900.Ciller C, De Zanet S, Kamnitsas K , et al. Multi-channel MRI segmentation of eye structures and tumors using patient-specific features[ J]. PLoS One, 2017, 12(3): e0173900.
47、Strijbis VIJ, de Bloeme CM, Jansen RW, et al. Multi-view convolutional neural networks for automated ocular structure and tumor segmentation in retinoblastoma[ J]. Sci Rep, 2021, 11(1): 14590.Strijbis VIJ, de Bloeme CM, Jansen RW, et al. Multi-view convolutional neural networks for automated ocular structure and tumor segmentation in retinoblastoma[ J]. Sci Rep, 2021, 11(1): 14590.
48、Goyal P, Padhi TR, Das T, et al. Outcome of universal newborn eye screening with wide-field digital retinal image acquisition system: a pilot study[ J]. Eye (Lond), 2018, 32(1): 67-73.Goyal P, Padhi TR, Das T, et al. Outcome of universal newborn eye screening with wide-field digital retinal image acquisition system: a pilot study[ J]. Eye (Lond), 2018, 32(1): 67-73.
49、Wang B, Xiao L, Liu Y, et al. Application of a deep convolutional neural network in the diagnosis of neonatal ocular fundus hemorrhage[ J]. Biosci Rep, 2018, 38(6): BSR20180497.Wang B, Xiao L, Liu Y, et al. Application of a deep convolutional neural network in the diagnosis of neonatal ocular fundus hemorrhage[ J]. Biosci Rep, 2018, 38(6): BSR20180497.
50、Hughes LA, May K, Talbot JF, et al. Incidence, distribution, and duration of birth-related retinal hemorrhages: a prospective study[ J]. J AAPOS, 2006, 10(2): 102-106.Hughes LA, May K, Talbot JF, et al. Incidence, distribution, and duration of birth-related retinal hemorrhages: a prospective study[ J]. J AAPOS, 2006, 10(2): 102-106.
51、Mao J, Luo Y, Chen K , et al. New grading criterion for retinal haemorrhages in term newborns based on deep convolutional neural networks[ J]. Clin Exp Ophthalmol, 2020, 48(2): 220-229.Mao J, Luo Y, Chen K , et al. New grading criterion for retinal haemorrhages in term newborns based on deep convolutional neural networks[ J]. Clin Exp Ophthalmol, 2020, 48(2): 220-229.
52、Xie Y, Sun L, Chen Y, et al. Ocular toxocariasis presenting as leukocoria[ J]. Lancet Infect Dis, 2022, 22(3): 426.Xie Y, Sun L, Chen Y, et al. Ocular toxocariasis presenting as leukocoria[ J]. Lancet Infect Dis, 2022, 22(3): 426.
53、Munson MC, Plewman DL, Baumer KM, et al. Autonomous early detection of eye disease in childhood photographs[ J]. Sci Adv, 2019, 5(10): eaax6363.Munson MC, Plewman DL, Baumer KM, et al. Autonomous early detection of eye disease in childhood photographs[ J]. Sci Adv, 2019, 5(10): eaax6363.
54、Zhang X, Shi YY. Prediction of myopic shift in paediatric pseudophakia using a neural network: a preliminary study[ J]. Chin J Ophthalmol, 2007, 43(11): 987-995.Zhang X, Shi YY. Prediction of myopic shift in paediatric pseudophakia using a neural network: a preliminary study[ J]. Chin J Ophthalmol, 2007, 43(11): 987-995.
55、Vogelsang L, Gilad-Gutnick S, Ehrenberg E, et al. Potential downside of high initial visual acuity[ J]. Proc Natl Acad Sci U S A, 2018, 115(44): 11333-11338.Vogelsang L, Gilad-Gutnick S, Ehrenberg E, et al. Potential downside of high initial visual acuity[ J]. Proc Natl Acad Sci U S A, 2018, 115(44): 11333-11338.
56、Grady CL, Mondloch CJ, Lewis TL, et al. Early visual deprivation from congenital cataracts disrupts activity and functional connectivity in the face network[ J]. Neuropsychologia, 2014, 57: 122-139.Grady CL, Mondloch CJ, Lewis TL, et al. Early visual deprivation from congenital cataracts disrupts activity and functional connectivity in the face network[ J]. Neuropsychologia, 2014, 57: 122-139.
57、Deshmukh AV, Badakere A, Sheth J, et al. Pivoting to teleconsultation for paediatric ophthalmology and strabismus: Our experience during COVID-19 times[ J]. Indian J Ophthalmol, 2020, 68(7): 1387-1391.Deshmukh AV, Badakere A, Sheth J, et al. Pivoting to teleconsultation for paediatric ophthalmology and strabismus: Our experience during COVID-19 times[ J]. Indian J Ophthalmol, 2020, 68(7): 1387-1391.
58、Guo Z, Ma N, Wu Y, et al. The safety and feasibility of the screening for retinopathy of prematurity assisted by telemedicine network during COVID-19 pandemic in Wuhan, China[ J]. BMC Ophthalmol, 2021, 21(1): 258.Guo Z, Ma N, Wu Y, et al. The safety and feasibility of the screening for retinopathy of prematurity assisted by telemedicine network during COVID-19 pandemic in Wuhan, China[ J]. BMC Ophthalmol, 2021, 21(1): 258.
上一篇
下一篇
作者登录
作者注册
审稿登录
编辑登录
其他期刊

  • 眼科学报

    主管：中华人民共和国教育部
    主办：中山大学
    承办：中山大学中山眼科中心
    主编：林浩添
    主管：中华人民共和国教育部
    主办：中山大学
    浏览

  • Eye Science

    主管：中华人民共和国教育部
    主办：中山大学
    承办：中山大学中山眼科中心
    主编：林浩添
    主管：中华人民共和国教育部
    主办：中山大学
    浏览
最新资讯

一亿年进化出的晶状体有多重要？

How important is the lens evolved in 100 million years?

发布于: 2022-12-01
推荐阅读
出版者信息

中山大学

中山大学中山眼科中心

眼科学术平台公众号
目录
中山大学医学期刊联盟
关注微信公众号
首页   | 关于我们  | 联系我们 粤ICP备11021180
关注微信公众号
粤ICP备11021180