Joint Shantou International Eye Center of Shantou University, the Chinese University of Hong Kong, Shantou, Guangdong, China.
Network and Information Center, Shantou University, Shantou, Guangdong, China.
JAMA Netw Open. 2021 May 3;4(5):e218758. doi: 10.1001/jamanetworkopen.2021.8758.
A retinopathy of prematurity (ROP) diagnosis currently relies on indirect ophthalmoscopy assessed by experienced ophthalmologists. A deep learning algorithm based on retinal images may facilitate early detection and timely treatment of ROP to improve visual outcomes.
To develop a retinal image-based, multidimensional, automated, deep learning platform for ROP screening and validate its performance accuracy.
DESIGN, SETTING, AND PARTICIPANTS: A total of 14 108 eyes of 8652 preterm infants who received ROP screening from 4 centers from November 4, 2010, to November 14, 2019, were included, and a total of 52 249 retinal images were randomly split into training, validation, and test sets. Four main dimensional independent classifiers were developed, including image quality, any stage of ROP, intraocular hemorrhage, and preplus/plus disease. Referral-warranted ROP was automatically generated by integrating the results of 4 classifiers at the image, eye, and patient levels. DeepSHAP, a method based on DeepLIFT and Shapley values (solution concepts in cooperative game theory), was adopted as the heat map technology to explain the predictions. The performance of the platform was further validated as compared with that of the experienced ROP experts. Data were analyzed from February 12, 2020, to June 24, 2020.
A deep learning algorithm.
The performance of each classifier included true negative, false positive, false negative, true positive, F1 score, sensitivity, specificity, receiver operating characteristic, area under curve (AUC), and Cohen unweighted κ.
A total of 14 108 eyes of 8652 preterm infants (mean [SD] gestational age, 32.9 [3.1] weeks; 4818 boys [60.4%] of 7973 with known sex) received ROP screening. The performance of all classifiers achieved an F1 score of 0.718 to 0.981, a sensitivity of 0.918 to 0.982, a specificity of 0.949 to 0.992, and an AUC of 0.983 to 0.998, whereas that of the referral system achieved an F1 score of 0.898 to 0.956, a sensitivity of 0.981 to 0.986, a specificity of 0.939 to 0.974, and an AUC of 0.9901 to 0.9956. Fine-grained and class-discriminative heat maps were generated by DeepSHAP in real time. The platform achieved a Cohen unweighted κ of 0.86 to 0.98 compared with a Cohen κ of 0.93 to 0.98 by the ROP experts.
In this diagnostic study, an automated ROP screening platform was able to identify and classify multidimensional pathologic lesions in the retinal images. This platform may be able to assist routine ROP screening in general and children hospitals.
目前，早产儿视网膜病变（ROP）的诊断依赖于经验丰富的眼科医生进行间接检眼镜评估。基于视网膜图像的深度学习算法可以帮助早期发现和及时治疗 ROP，从而改善视觉预后。
开发一种基于视网膜图像的多维、自动化的深度学习平台，用于 ROP 筛查，并验证其性能准确性。
设计、设置和参与者：共纳入了来自 4 个中心的 108 只眼 8652 例早产儿在 2010 年 11 月 4 日至 2019 年 11 月 14 日期间接受 ROP 筛查的数据，共采集了 52249 张视网膜图像，随机分为训练集、验证集和测试集。开发了 4 个主要的多维独立分类器，包括图像质量、任何阶段的 ROP、眼内出血和 Preplus/plus 疾病。通过整合 4 个分类器在图像、眼和患者层面的结果，自动生成需要转诊的 ROP。采用 DeepSHAP（基于 DeepLIFT 和 Shapley 值的方法）作为热图技术来解释预测。进一步与有经验的 ROP 专家的表现进行比较，验证了该平台的性能。数据分析于 2020 年 2 月 12 日至 6 月 24 日进行。
深度学习算法。
每个分类器的性能包括真阴性、假阳性、假阴性、真阳性、F1 评分、敏感性、特异性、接受者操作特征、曲线下面积（AUC）和 Cohen 未加权κ。
共纳入 108 只眼 8652 例早产儿（平均[标准差]胎龄为 32.9[3.1]周；已知性别 7973 例中男孩 4818 例[60.4%]）接受 ROP 筛查。所有分类器的 F1 评分均为 0.718 至 0.981，敏感性为 0.918 至 0.982，特异性为 0.949 至 0.992，AUC 为 0.983 至 0.998，而转诊系统的 F1 评分为 0.898 至 0.956，敏感性为 0.981 至 0.986，特异性为 0.939 至 0.974，AUC 为 0.9901 至 0.9956。DeepSHAP 实时生成了细粒度和类区分的热图。与 ROP 专家的 Cohen κ 值为 0.93 至 0.98 相比，该平台的 Cohen 未加权κ 值为 0.86 至 0.98。
在这项诊断研究中，一种自动化的 ROP 筛查平台能够识别和分类视网膜图像中的多维病理病变。该平台有望辅助常规 ROP 筛查，并在综合医院和儿童医院中得到应用。
