当前位置: X-MOL 学术N. Engl. J. Med. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Artificial Intelligence to Detect Papilledema from Ocular Fundus Photographs.
The New England Journal of Medicine ( IF 158.5 ) Pub Date : 2020-04-14 , DOI: 10.1056/nejmoa1917130
Dan Milea 1 , Raymond P Najjar 1 , Jiang Zhubo 1 , Daniel Ting 1 , Caroline Vasseneix 1 , Xinxing Xu 1 , Masoud Aghsaei Fard 1 , Pedro Fonseca 1 , Kavin Vanikieti 1 , Wolf A Lagrèze 1 , Chiara La Morgia 1 , Carol Y Cheung 1 , Steffen Hamann 1 , Christophe Chiquet 1 , Nicolae Sanda 1 , Hui Yang 1 , Luis J Mejico 1 , Marie-Bénédicte Rougier 1 , Richard Kho 1 , Tran Thi Ha Chau 1 , Shweta Singhal 1 , Philippe Gohier 1 , Catherine Clermont-Vignal 1 , Ching-Yu Cheng 1 , Jost B Jonas 1 , Patrick Yu-Wai-Man 1 , Clare L Fraser 1 , John J Chen 1 , Selvakumar Ambika 1 , Neil R Miller 1 , Yong Liu 1 , Nancy J Newman 1 , Tien Y Wong 1 , Valérie Biousse 1 ,
Affiliation  

BACKGROUND Nonophthalmologist physicians do not confidently perform direct ophthalmoscopy. The use of artificial intelligence to detect papilledema and other optic-disk abnormalities from fundus photographs has not been well studied. METHODS We trained, validated, and externally tested a deep-learning system to classify optic disks as being normal or having papilledema or other abnormalities from 15,846 retrospectively collected ocular fundus photographs that had been obtained with pharmacologic pupillary dilation and various digital cameras in persons from multiple ethnic populations. Of these photographs, 14,341 from 19 sites in 11 countries were used for training and validation, and 1505 photographs from 5 other sites were used for external testing. Performance at classifying the optic-disk appearance was evaluated by calculating the area under the receiver-operating-characteristic curve (AUC), sensitivity, and specificity, as compared with a reference standard of clinical diagnoses by neuro-ophthalmologists. RESULTS The training and validation data sets from 6779 patients included 14,341 photographs: 9156 of normal disks, 2148 of disks with papilledema, and 3037 of disks with other abnormalities. The percentage classified as being normal ranged across sites from 9.8 to 100%; the percentage classified as having papilledema ranged across sites from zero to 59.5%. In the validation set, the system discriminated disks with papilledema from normal disks and disks with nonpapilledema abnormalities with an AUC of 0.99 (95% confidence interval [CI], 0.98 to 0.99) and normal from abnormal disks with an AUC of 0.99 (95% CI, 0.99 to 0.99). In the external-testing data set of 1505 photographs, the system had an AUC for the detection of papilledema of 0.96 (95% CI, 0.95 to 0.97), a sensitivity of 96.4% (95% CI, 93.9 to 98.3), and a specificity of 84.7% (95% CI, 82.3 to 87.1). CONCLUSIONS A deep-learning system using fundus photographs with pharmacologically dilated pupils differentiated among optic disks with papilledema, normal disks, and disks with nonpapilledema abnormalities. (Funded by the Singapore National Medical Research Council and the SingHealth Duke-NUS Ophthalmology and Visual Sciences Academic Clinical Program.).

中文翻译:

从眼底照片中检测视乳头水肿的人工智能。

背景非眼科医师不能自信地进行直接检眼镜检查。使用人工智能从眼底照片中检测视乳头水肿和其他视盘异常尚未得到很好的研究。方法 我们训练、验证和外部测试了一个深度学习系统,以根据 15,846 张回顾性收集的眼底照片将视盘分类为正常或有视神经乳头水肿或其他异常,这些照片是通过药理瞳孔扩张和各种数码相机从多个人群中获得的。种族人口。在这些照片中,来自 11 个国家/地区的 19 个站点的 14,341 张用于训练和验证,来自其他 5 个站点的 1505 张照片用于外部测试。与神经眼科医生的临床诊断参考标准相比,通过计算接受者操作特征曲线 (AUC) 下的面积、灵敏度和特异性来评估对视盘外观进行分类的性能。结果 来自 6779 名患者的训练和验证数据集包括 14,341 张照片:9156 张正常椎间盘、2148 张视乳头水肿椎间盘和 3037 张其他异常椎间盘。被归类为正常的百分比在各个站点之间介于 9.8% 和 100% 之间;被归类为视乳头水肿的百分比范围从零到 59.5%。在验证集中,系统将具有视乳头水肿的视盘与正常视盘和具有非视乳头水肿异常的视盘区分开来,AUC 为 0.99(95% 置信区间 [CI],0.98 至 0. 99) 和异常磁盘的正常值,AUC 为 0.99(95% CI,0.99 至 0.99)。在 1505 张照片的外部测试数据集中,系统检测视乳头水肿的 AUC 为 0.96(95% CI,0.95 至 0.97),灵敏度为 96.4%(95% CI,93.9 至 98.3),并且特异性为 84.7%(95% CI,82.3 至 87.1)。结论 一个深度学习系统使用具有药理散大瞳孔的眼底照片来区分具有视乳头水肿的视盘、正常视盘和具有非视乳头水肿异常的视盘。(由新加坡国家医学研究委员会和 SingHealth Duke-NUS 眼科和视觉科学学术临床项目资助。)。灵敏度为 96.4%(95% CI,93.9 至 98.3),特异性为 84.7%(95% CI,82.3 至 87.1)。结论 一个深度学习系统使用具有药理散大瞳孔的眼底照片来区分具有视乳头水肿的视盘、正常视盘和具有非视乳头水肿异常的视盘。(由新加坡国家医学研究委员会和 SingHealth Duke-NUS 眼科和视觉科学学术临床项目资助。)。灵敏度为 96.4%(95% CI,93.9 至 98.3),特异性为 84.7%(95% CI,82.3 至 87.1)。结论 一个深度学习系统使用具有药理散大瞳孔的眼底照片来区分具有视乳头水肿的视盘、正常视盘和具有非视乳头水肿异常的视盘。(由新加坡国家医学研究委员会和 SingHealth Duke-NUS 眼科和视觉科学学术临床项目资助。)。
更新日期:2020-04-14
down
wechat
bug