Department of Ophthalmology (TYAL, NRM), Wilmer Eye Institute, Johns Hopkins University, Baltimore, Maryland; Department of Ophthalmology (DSWT), Singapore Eye Research Institute, Singapore National Eye Center, Duke-NUS Medical School, National University of Singapore, Singapore; Department of Radiology (PHY, FKH), Johns Hopkins University, Baltimore, Maryland; Department of Biomedical Engineering (JW), Johns Hopkins University, Baltimore, Maryland; Computational Interaction and Robotics Lab (HZ, GDH), Johns Hopkins University, Baltimore, Maryland; Department of Ophthalmology (PSS), University of Colorado School of Medicine, Aurora, Colorado; School of Medicine (TL), Johns Hopkins University, Baltimore, Maryland; and Malone Center for Engineering in Healthcare (GDH), Johns Hopkins University, Baltimore, Maryland.
J Neuroophthalmol. 2020 Jun;40(2):178-184. doi: 10.1097/WNO.0000000000000827.
Deep learning (DL) has demonstrated human expert levels of performance for medical image classification in a wide array of medical fields, including ophthalmology. In this article, we present the results of our DL system designed to determine optic disc laterality, right eye vs left eye, in the presence of both normal and abnormal optic discs.
Using transfer learning, we modified the ResNet-152 deep convolutional neural network (DCNN), pretrained on ImageNet, to determine the optic disc laterality. After a 5-fold cross-validation, we generated receiver operating characteristic curves and corresponding area under the curve (AUC) values to evaluate performance. The data set consisted of 576 color fundus photographs (51% right and 49% left). Both 30° photographs centered on the optic disc (63%) and photographs with varying degree of optic disc centration and/or wider field of view (37%) were included. Both normal (27%) and abnormal (73%) optic discs were included. Various neuro-ophthalmological diseases were represented, such as, but not limited to, atrophy, anterior ischemic optic neuropathy, hypoplasia, and papilledema.
Using 5-fold cross-validation (70% training; 10% validation; 20% testing), our DCNN for classifying right vs left optic disc achieved an average AUC of 0.999 (±0.002) with optimal threshold values, yielding an average accuracy of 98.78% (±1.52%), sensitivity of 98.60% (±1.72%), and specificity of 98.97% (±1.38%). When tested against a separate data set for external validation, our 5-fold cross-validation model achieved the following average performance: AUC 0.996 (±0.005), accuracy 97.2% (±2.0%), sensitivity 96.4% (±4.3%), and specificity 98.0% (±2.2%).
Small data sets can be used to develop high-performing DL systems for semantic labeling of neuro-ophthalmology images, specifically in distinguishing between right and left optic discs, even in the presence of neuro-ophthalmological pathologies. Although this may seem like an elementary task, this study demonstrates the power of transfer learning and provides an example of a DCNN that can help curate large medical image databases for machine-learning purposes and facilitate ophthalmologist workflow by automatically labeling images according to laterality.
深度学习（DL）已经在包括眼科在内的广泛医学领域中展示了对医学图像分类的专家级表现。在本文中，我们展示了我们设计的用于确定正常和异常视盘存在时的视盘侧别（右眼与左眼）的 DL 系统的结果。
使用迁移学习，我们修改了在 ImageNet 上进行预训练的 ResNet-152 深度卷积神经网络（DCNN），以确定视盘侧别。在进行 5 折交叉验证后，我们生成了接收器工作特性曲线和相应的曲线下面积（AUC）值来评估性能。数据集包含 576 张彩色眼底照片（51%为右眼，49%为左眼）。包括 30°以视盘为中心的照片（63%）和视盘偏心度和/或视野更宽的照片（37%）。包括正常（27%）和异常（73%）视盘。代表了各种神经眼科疾病，例如但不限于萎缩、前部缺血性视神经病变、发育不良和视乳头水肿。
使用 5 折交叉验证（70%训练；10%验证；20%测试），我们用于分类右眼与左眼视盘的 DCNN 的平均 AUC 为 0.999（±0.002），最佳阈值值，平均准确率为 98.78%（±1.52%），灵敏度为 98.60%（±1.72%），特异性为 98.97%（±1.38%）。当在外部验证的单独数据集上进行测试时，我们的 5 折交叉验证模型的平均性能为：AUC 0.996（±0.005）、准确率 97.2%（±2.0%）、灵敏度 96.4%（±4.3%）和特异性 98.0%（±2.2%）。
即使存在神经眼科病理学，也可以使用小数据集开发用于神经眼科图像语义标记的高性能 DL 系统，特别是用于区分右眼和左眼视盘。虽然这似乎是一项基本任务，但本研究展示了迁移学习的强大功能，并提供了一个示例，说明 DCNN 可以帮助为机器学习目的整理大型医学图像数据库，并通过根据侧别自动标记图像来简化眼科医生的工作流程。
