From the NIHR Biomedical Research Centre at Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology (G.L., G.M., J.M.-N., D.F.G.-H.), London, United Kingdom; Centre for Medical Image Computing, University College London (G.L.), London, United Kingdom.
From the NIHR Biomedical Research Centre at Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology (G.L., G.M., J.M.-N., D.F.G.-H.), London, United Kingdom; Optometry and Visual Sciences, City, University of London, London, United Kingdom.
Am J Ophthalmol. 2022 Jun;238:52-65. doi: 10.1016/j.ajo.2021.12.020. Epub 2022 Jan 5.
To develop and validate a deep learning method of predicting visual function from spectral domain optical coherence tomography (SD-OCT)-derived retinal nerve fiber layer thickness (RNFLT) measurements and corresponding SD-OCT images.
Development and evaluation of diagnostic technology.
Two deep learning ensemble models to predict pointwise VF sensitivity from SD-OCT images (model 1: RNFLT profile only; model 2: RNFLT profile plus SD-OCT image) and 2 reference models were developed. All models were tested in an independent test-retest data set comprising 2181 SD-OCT/VF pairs; the median of ∼10 VFs per eye was taken as the best available estimate (BAE) of the true VF. The performance of single VFs predicting the BAE VF was also evaluated. The training data set comprised 954 eyes of 220 healthy and 332 glaucomatous participants, and the test data set, 144 eyes of 72 glaucomatous participants. The main outcome measures included the pointwise prediction mean error (ME), mean absolute error (MAE), and correlation of predictions with the BAE VF sensitivity.
The median mean deviation was -4.17 dB (-14.22 to 0.88). Model 2 had excellent accuracy (ME 0.5 dB, SD 0.8) and overall performance (MAE 2.3 dB, SD 3.1), and significantly (paired t test) outperformed the other methods. For single VFs predicting the BAE VF, the pointwise MAE was 1.5 dB (SD 0.7). The association between SD-OCT and single VF predictions of the BAE pointwise VF sensitivities was R = 0.78 and R = 0.88, respectively.
Our method outperformed standard statistical and deep learning approaches. Predictions of BAEs from OCT images approached the accuracy of single real VF estimates of the BAE.
开发并验证一种基于谱域光学相干断层扫描（SD-OCT）衍生的视网膜神经纤维层厚度（RNFLT）测量值和相应的 SD-OCT 图像预测视功能的深度学习方法。
诊断技术的开发和评估。
开发了两种基于深度学习的集合模型，用于从 SD-OCT 图像预测点状 VF 敏感性（模型 1：仅 RNFLT 谱；模型 2：RNFLT 谱加 SD-OCT 图像）和 2 个参考模型。所有模型均在包含 2181 对 SD-OCT/VF 的独立测试-复测数据集中进行了测试；每只眼的中位数约为 10 个 VF，作为真实 VF 的最佳可用估计值（BAE）。还评估了单个 VF 预测 BAE VF 的性能。训练数据集包含 220 名健康人和 332 名青光眼患者的 954 只眼，测试数据集包含 72 名青光眼患者的 144 只眼。主要观察指标包括点预测平均误差（ME）、平均绝对误差（MAE）和预测值与 BAE VF 敏感性的相关性。
中位平均偏差为-4.17dB（-14.22 至 0.88）。模型 2具有出色的准确性（ME 0.5dB，SD 0.8）和整体性能（MAE 2.3dB，SD 3.1），并显著（配对 t 检验）优于其他方法。对于单个 VF 预测 BAE VF，点预测 MAE 为 1.5dB（SD 0.7）。SD-OCT 与 BAE 点预测 VF 敏感性的单个 VF 预测之间的关联分别为 R=0.78 和 R=0.88。
我们的方法优于标准统计和深度学习方法。从 OCT 图像预测 BAE 接近单个真实 VF 对 BAE 估计值的准确性。
