Wang Yining, Chen Changyu, Wang Ziye, Wu Yijin, Lu Hongshuang, Xiong Jianping, Sugisawa Keigo, Kamoi Koju, Ohno-Matsui Kyoko
Department of Ophthalmology and Visual Science, Institute of Science Tokyo, Tokyo, Japan.
Transl Vis Sci Technol. 2025 Jun 2;14(6):25. doi: 10.1167/tvst.14.6.25.
To develop a deep learning (DL) model for screening posterior staphylomas in highly myopic patients using ultra-widefield optical coherence tomography (UWF-OCT) images.
Our retrospective single-center study collected 1428 qualified UWF-OCT images from 438 highly myopic patients between 2017 and 2019 for model development. An independent test dataset for internal validation included 216 images from 69 highly myopic patients obtained between June 2020 and December 2020. Posterior staphylomas were detected by identifying the staphyloma edges. Seven independent architectures (VGG16, VGG19, ResNet18, ResNet50, ResNet101, DenseNet121, and DenseNet161) were used to train the models and identify staphyloma edges. The area under the receiver operating characteristic (ROC) curve (AUC) was used to evaluate and compare the performance of each model.
The AUCs of seven DL models ranged from 0.794 (95% confidence interval [CI], 0.708-0.875) to 0.903 (95% CI, 0.846-0.953) for staphyloma edge detection in the internal test dataset. VGG19, with the highest AUC, achieved sensitivity (0.871; 95% CI, 0.773-0.931) that was comparable to or better than those of retina specialists. Heatmaps showed that the DL models could precisely identify the region of staphyloma edges.
Our models reliably identified staphyloma edges with high sensitivity and specificity. Given that posterior staphylomas are a key contributor to various fundus complications, the development of DL models holds significant promise for improving the clinical management of highly myopic patients.
This effective artificial intelligence system can help ophthalmologists screen posterior staphylomas in highly myopic eyes.
利用超广角光学相干断层扫描（UWF-OCT）图像开发一种深度学习（DL）模型，用于筛查高度近视患者的后巩膜葡萄肿。
我们的回顾性单中心研究收集了2017年至2019年间438例高度近视患者的1428张合格UWF-OCT图像用于模型开发。用于内部验证的独立测试数据集包括2020年6月至2020年12月间69例高度近视患者的216张图像。通过识别葡萄肿边缘来检测后巩膜葡萄肿。使用七种独立架构（VGG16、VGG19、ResNet18、ResNet50、ResNet101、DenseNet121和DenseNet161）训练模型并识别葡萄肿边缘。采用受试者操作特征（ROC）曲线下面积（AUC）评估和比较各模型的性能。
在内部测试数据集中，七种DL模型用于葡萄肿边缘检测的AUC范围为0.794（95%置信区间[CI]，0.708 - 0.875）至0.903（95%CI，0.846 - 0.953）。AUC最高的VGG模型的灵敏度（0.871；95%CI，0.773 - 0.931）与视网膜专家相当或更好。热图显示DL模型能够精确识别葡萄肿边缘区域。
我们的模型能够以高灵敏度和特异性可靠地识别葡萄肿边缘。鉴于后巩膜葡萄肿是各种眼底并发症的关键因素，DL模型 的开发对于改善高度近视患者的临床管理具有重要意义。
这种有效的人工智能系统可帮助眼科医生筛查高度近视眼中的后巩膜葡萄肿。
