Department of Ophthalmology, Xinhua Hospital Affiliated to Shanghai Jiaotong University School of Medicine, Shanghai, China.
Joint Shantou International Eye Center of Shantou University and the Chinese University of Hong Kong, Shantou University Medical College, Shantou, Guangdong, China.
Transl Vis Sci Technol. 2021 Apr 1;10(4):34. doi: 10.1167/tvst.10.4.34.
To develop generative adversarial networks (GANs) that synthesize realistic anterior segment optical coherence tomography (AS-OCT) images and evaluate deep learning (DL) models that are trained on real and synthetic datasets for detecting angle closure.
The GAN architecture was adopted and trained on the dataset with AS-OCT images collected from the Joint Shantou International Eye Center of Shantou University and the Chinese University of Hong Kong, synthesizing open- and closed-angle AS-OCT images. A visual Turing test with two glaucoma specialists was performed to assess the image quality of real and synthetic images. DL models, trained on either real or synthetic datasets, were developed. Using the clinicians' grading of the AS-OCT images as the reference standard, we compared the diagnostic performance of open-angle vs. closed-angle detection of DL models and the AS-OCT parameter, defined as a trabecular-iris space area 750 µm anterior to the scleral spur (TISA750), in a small independent validation dataset.
The GAN training included 28,643 AS-OCT anterior chamber angle (ACA) images. The real and synthetic datasets for DL model training have an equal distribution of open- and closed-angle images (all with 10,000 images each). The independent validation dataset included 238 open-angle and 243 closed-angle AS-OCT ACA images. The image quality of real versus synthetic AS-OCT images was similar, as assessed by the two glaucoma specialists, except for the scleral spur visibility. For the independent validation dataset, both DL models achieved higher areas under the curve compared with TISA750. Two DL models had areas under the curve of 0.97 (95% confidence interval, 0.96-0.99) and 0.94 (95% confidence interval, 0.92-0.96).
The GAN synthetic AS-OCT images appeared to be of good quality, according to the glaucoma specialists. The DL models, trained on all-synthetic AS-OCT images, can achieve high diagnostic performance.
The GANs can generate realistic AS-OCT images, which can also be used to train DL models.
开发生成对抗网络（GAN），合成逼真的眼前节光学相干断层扫描（AS-OCT）图像，并评估基于真实和合成数据集训练的深度学习（DL）模型，以检测房角关闭。
采用 GAN 架构，对来自汕头大学和香港中文大学联合汕头国际眼科中心的 AS-OCT 图像数据集进行训练，合成开角和闭角 AS-OCT 图像。通过两位青光眼专家进行视觉图灵测试，评估真实和合成图像的图像质量。开发了基于真实或合成数据集训练的 DL 模型。使用临床医生对 AS-OCT 图像的分级作为参考标准，我们比较了 DL 模型对开角和闭角检测的诊断性能，以及作为巩膜突前 750 µm 处小梁虹膜空间面积（TISA750）的 AS-OCT 参数在一个小型独立验证数据集的诊断性能。
GAN 训练包括 28643 张 AS-OCT 眼前房角（ACA）图像。用于 DL 模型训练的真实和合成数据集具有相同的开角和闭角图像分布（各有 10000 张图像）。独立验证数据集包括 238 张开角和 243 张闭角 AS-OCT ACA 图像。两位青光眼专家评估发现，真实与合成 AS-OCT 图像的图像质量相似，除了巩膜突可见性。对于独立验证数据集，与 TISA750 相比，两种 DL 模型的曲线下面积均更高。两个 DL 模型的曲线下面积分别为 0.97（95%置信区间，0.96-0.99）和 0.94（95%置信区间，0.92-0.96）。
根据青光眼专家的评估，GAN 合成的 AS-OCT 图像质量似乎较好。基于所有合成 AS-OCT 图像训练的 DL 模型可以实现较高的诊断性能。
翻译是否准确流畅？请根据你的理解进行调整。
