Department of Ophthalmology, Yeungnam University College of Medicine, Daegu, South Korea.
Nune Eye Hospital, Daegu, South Korea.
Transl Vis Sci Technol. 2022 Feb 1;11(2):39. doi: 10.1167/tvst.11.2.39.
To develop an automated diabetic retinopathy (DR) staging system using optical coherence tomography angiography (OCTA) images with a convolutional neural network (CNN) and to verify the feasibility of the system.
In this retrospective cross-sectional study, a total of 918 data sets of 3 × 3 mm2 OCTA images and 917 data sets of 6 × 6 mm2 OCTA images were obtained from 1118 eyes. A deep CNN and four traditional machine learning models were trained with annotations made by a retinal specialist based on ultra-widefield fluorescein angiography. Separately, the same images of the test data sets were independently graded by two human experts. The results of the CNN algorithm were compared with those of traditional machine learning-based classifiers and human experts.
The proposed CNN achieved an accuracy of 0.728, a sensitivity of 0.675, a specificity of 0.944, an F1 score of 0.683, and a quadratic weighted κ of 0.908 for a six-level staging task, which were far superior to the results of traditional machine learning methods or human experts. The CNN algorithm showed a better performance using 6 × 6 mm2 rather than 3 × 3 mm2 sized OCTA images and using combined data rather than a separate OCTA layer alone.
CNN-based classification using OCTA images can provide reliable assistance to clinicians for DR classification.
This CNN algorithm can guide the clinical decision for invasive angiography or referrals to ophthalmology specialists, helping to create more efficient diagnostic workflow in primary care settings.
使用卷积神经网络（CNN）开发一种基于光相干断层扫描血管造影（OCTA）图像的自动化糖尿病视网膜病变（DR）分期系统，并验证该系统的可行性。
在这项回顾性的横断面研究中，共获得了 1118 只眼中的 918 个 3×3mm²OCTA 图像数据集和 917 个 6×6mm²OCTA 图像数据集。基于超广角荧光素血管造影，由一名视网膜专家对 OCTA 图像进行注释，利用深度 CNN 和四种传统机器学习模型对这些数据进行训练。然后，使用两个独立的人类专家对测试数据集的相同图像进行独立分级。将 CNN 算法的结果与传统基于机器学习的分类器和人类专家的结果进行比较。
提出的 CNN 算法在六级分期任务中的准确率为 0.728，灵敏度为 0.675，特异性为 0.944，F1 得分为 0.683，二次加权κ值为 0.908，远优于传统机器学习方法或人类专家的结果。与使用 3×3mm²大小的 OCTA 图像或单独 OCTA 层相比，使用 6×6mm²大小的 OCTA 图像和使用组合数据的 CNN 算法具有更好的性能。
基于 OCTA 图像的 CNN 分类可以为 DR 分类提供可靠的临床辅助。
马萨诸塞州眼耳医院，波士顿，美国。
