School of Electronic Engineering and Computer Science, Queen Mary University of London, London, UK.
Center for Rehabilitation Medicine, Department of Ophthalmology, Zhejiang Provincial People's Hospital (Affiliated People's Hospital, Hangzhou Medical College), Hangzhou, Zhejiang, China.
Med Phys. 2023 Sep;50(9):5398-5409. doi: 10.1002/mp.16623. Epub 2023 Jul 25.
Myopic traction maculopathy (MTM) are retinal disorder caused by traction force on the macula, which can lead to varying degrees of vision loss in eyes with high myopia. Optical coherence tomography (OCT) is an effective imaging technique for diagnosing, detecting and classifying retinopathy. MTM has been classified into different patterns by OCT, corresponding to different clinical strategies.
We aimed to engineer a deep learning model that can automatically identify MTM in highly myopic (HM) eyes using OCT images.
A five-class classification model was developed using 2837 OCT images from 958 HM patients. We adopted a ResNet-34 architecture to train the model to identify MTM: no MTM (class 0), extra-foveal maculoschisis (class 1), inner lamellar macular hole (class 2), outer foveoschisis (class 3), and discontinuity or detachment of foveal outer hyperreflective layers (class 4). An independent test set of 604 images from 173 HM patients was used to evaluate the model's performance. Classification performance was assessed according to the area under the curve (AUC), accuracy, sensitivity, specificity.
Our model exhibited a high training performance for classification (F1-score of 0.953; AUCs of 0.961 to 0.998). In test set, it achieved sensitivities (91.67%-97.78 %) and specificities (98.33%-99.17%) as good as, or better than, those of experienced clinicians. Heatmaps were generated to provide visual explanations.
We established a deep learning model for MTM classification using OCT images. This model performed equally well or better than retinal specialists and is suitable for large-scale screening and identifying MTM in HM eyes.
近视牵引性黄斑病变（MTM）是一种黄斑区受牵引导致的视网膜病变，可导致高度近视眼视力不同程度下降。光学相干断层扫描（OCT）是一种有效的成像技术，可用于诊断、检测和分类视网膜病变。MTM 可根据 OCT 分为不同类型，对应不同的临床策略。
我们旨在开发一种深度学习模型，利用 OCT 图像自动识别高度近视（HM）眼中的 MTM。
采用 ResNet-34 架构，利用 958 例 HM 患者的 2837 张 OCT 图像，建立五分类模型，识别 MTM：无 MTM（0 类）、黄斑区神经上皮层劈裂（1 类）、黄斑内层裂孔（2 类）、黄斑外劈裂（3 类）、黄斑外层高反射带连续性或脱离（4 类）。采用 173 例 HM 患者的 604 张独立测试集 OCT 图像评估模型性能。根据曲线下面积（AUC）、准确性、敏感性、特异性评估分类性能。
模型对分类具有较高的训练性能（F1 评分 0.953；AUC 为 0.961-0.998）。在测试集中，其敏感性（91.67%-97.78%）和特异性（98.33%-99.17%）与经验丰富的临床医生相当，甚至更好。生成热图提供可视化解释。
我们建立了一种基于 OCT 图像的 MTM 分类深度学习模型。该模型与视网膜专家表现相当或更好，适用于大规模筛查和识别 HM 眼中的 MTM。
