Department of Ophthalmology, Tsukazaki Hospital, Himeji, Japan.
Rist Incorporated, Tokyo, Japan.
PLoS One. 2019 Nov 7;14(11):e0223965. doi: 10.1371/journal.pone.0223965. eCollection 2019.
We aimed to assess the ability of deep learning (DL) and support vector machine (SVM) to detect a nonperfusion area (NPA) caused by retinal vein occlusion (RVO) with optical coherence tomography angiography (OCTA) images. The study included 322 OCTA images (normal: 148; NPA owing to RVO: 174 [128 branch RVO images and 46 central RVO images]). Training to construct the DL model using deep convolutional neural network (DNN) algorithms was provided using OCTA images. The SVM used a scikit-learn library with a radial basis function kernel. The area under the curve (AUC), sensitivity and specificity for detecting an NPA were examined. We compared the diagnostic ability (sensitivity, specificity and average required time) between the DNN, SVM and seven ophthalmologists. Heat maps were generated. With regard to the DNN, the mean AUC, sensitivity, specificity and average required time for distinguishing RVO OCTA images with an NPA from normal OCTA images were 0.986, 93.7%, 97.3% and 176.9 s, respectively. With regard to SVM, the mean AUC, sensitivity, and specificity were 0.880, 79.3%, and 81.1%, respectively. With regard to the seven ophthalmologists, the mean AUC, sensitivity, specificity and average required time were 0.962, 90.8%, 89.2%, and 700.6 s, respectively. The DNN focused on the foveal avascular zone and NPA in heat maps. The performance of the DNN was significantly better than that of SVM in all parameters (p < 0.01, all) and that of the ophthalmologists in AUC and specificity (p < 0.01, all). The combination of DL and OCTA images had high accuracy for the detection of an NPA, and it might be useful in clinical practice and retinal screening.
我们旨在评估深度学习 (DL) 和支持向量机 (SVM) 利用光相干断层扫描血管造影 (OCTA) 图像检测视网膜静脉阻塞 (RVO) 引起的无灌注区 (NPA) 的能力。该研究包括 322 张 OCTA 图像（正常：148 张；NPA 归因于 RVO：174 张[128 张分支 RVO 图像和 46 张中央 RVO 图像]）。使用 OCTA 图像为使用深度卷积神经网络 (DNN) 算法构建 DL 模型提供训练。SVM 使用带有径向基函数核的 scikit-learn 库。检查了用于检测 NPA 的曲线下面积 (AUC)、敏感性和特异性。我们比较了 DNN、SVM 和七名眼科医生的诊断能力（敏感性、特异性和平均所需时间）。生成了热图。对于 DNN，用于区分具有 NPA 的 RVO OCTA 图像和正常 OCTA 图像的平均 AUC、敏感性、特异性和平均所需时间分别为 0.986、93.7%、97.3%和 176.9s。对于 SVM，平均 AUC、敏感性和特异性分别为 0.880、79.3%和 81.1%。对于七名眼科医生，平均 AUC、敏感性、特异性和平均所需时间分别为 0.962、90.8%、89.2%和 700.6s。DNN 专注于黄斑无血管区和热图中的 NPA。在所有参数（p < 0.01，全部）中，DNN 的性能明显优于 SVM，在 AUC 和特异性方面也明显优于眼科医生（p < 0.01，全部）。DL 和 OCTA 图像的组合对 NPA 的检测具有很高的准确性，可能对临床实践和视网膜筛查有用。
