From the Glaucoma Division, Stein Eye Institute, David Geffen School of Medicine, University of California Los Angeles (V.M., S.B., E.M., K.E., M.R., A.M., D.Z., J.C., K.N.-M.), Los Angeles, California, USA.
Department of Computer Science, Pepperdine University (S.W., F.S.), Malibu, California, USA.
Am J Ophthalmol. 2024 Jun;262:141-152. doi: 10.1016/j.ajo.2024.02.007. Epub 2024 Feb 12.
Identifying glaucoma patients at high risk of progression based on widely available structural data is an unmet task in clinical practice. We test the hypothesis that baseline or serial structural measures can predict visual field (VF) progression with deep learning (DL).
Development of a DL algorithm to predict VF progression.
3,079 eyes (1,765 patients) with various types of glaucoma and ≥5 VFs, and ≥3 years of follow-up from a tertiary academic center were included. Serial VF mean deviation (MD) rates of change were estimated with linear-regression. VF progression was defined as negative MD slope with p<0.05. A Siamese Neural Network with ResNet-152 backbone pre-trained on ImageNet was designed to predict VF progression using serial optic-disc photographs (ODP), and baseline retinal nerve fiber layer (RNFL) thickness. We tested the model on a separate dataset (427 eyes) with RNFL data from different OCT. The Main Outcome Measure was Area under ROC curve (AUC).
Baseline average (SD) MD was 3.4 (4.9)dB. VF progression was detected in 900 eyes (29%). AUC (95% CI) for model incorporating baseline ODP and RNFL thickness was 0.813 (0.757-0.869). After adding the second and third ODPs, AUC increased to 0.860 and 0.894, respectively (p<0.027). This model also had highest AUC (0.911) for predicting fast progression (MD rate <1.0 dB/year). Model's performance was similar when applied to second dataset using RNFL data from another OCT device (AUC=0.893; 0.837-0.948).
DL model predicted VF progression with clinically relevant accuracy using baseline RNFL thickness and serial ODPs and can be implemented as a clinical tool after further validation.
基于广泛可用的结构数据识别具有进展风险的青光眼患者是临床实践中的一项未满足的任务。我们检验了基于深度学习（DL）的基线或连续结构测量能否预测视野（VF）进展的假设。
开发一种用于预测 VF 进展的 DL 算法。
纳入了来自一家三级学术中心的各种类型青光眼患者的 3079 只眼（1765 例患者），这些患者至少有 5 次 VF 检查和≥3 年的随访。使用线性回归估计了连续 VF 平均偏差（MD）变化率。VF 进展定义为 MD 斜率为负且 p<0.05。设计了一个带有 ResNet-152 主干的孪生神经网络，该网络使用来自不同 OCT 的连续视盘照片（ODP）和基线视网膜神经纤维层（RNFL）厚度来预测 VF 进展。我们在一个具有来自不同 OCT 的 RNFL 数据的独立数据集（427 只眼）上测试了该模型。主要观察指标是 ROC 曲线下面积（AUC）。
基线平均（SD）MD 为 3.4（4.9）dB。900 只眼（29%）检测到 VF 进展。纳入基线 ODP 和 RNFL 厚度的模型 AUC（95%CI）为 0.813（0.757-0.869）。加入第二和第三张 ODP 后，AUC 分别增加到 0.860 和 0.894（p<0.027）。该模型在预测快速进展（MD 率<1.0dB/年）时也具有最高 AUC（0.911）。当将其应用于使用另一台 OCT 设备的 RNFL 数据的第二个数据集时，模型的性能相似（AUC=0.893；0.837-0.948）。
使用基线 RNFL 厚度和连续 ODP，DL 模型以具有临床相关性的准确性预测了 VF 进展，可以在进一步验证后作为临床工具实施。
