Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania.
Department of Ophthalmology, Center for Preventive Ophthalmology and Biostatistics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania.
Ophthalmol Retina. 2024 May;8(5):419-430. doi: 10.1016/j.oret.2023.11.010. Epub 2023 Nov 24.
To evaluate multiple machine learning (ML) models for predicting 2-year visual acuity (VA) responses to anti-vascular endothelial growth factor (anti-VEGF) treatment in the Comparison of Age-related Macular Degeneration (AMD) Treatments Trials (CATT) for patients with neovascular AMD (nAMD).
Secondary analysis of public data from a randomized clinical trial.
A total of 1029 CATT participants who completed 2 years of follow-up with untreated active nAMD and baseline VA between 20/25 and 20/320 in the study eye.
Five ML models (support vector machine, random forest, extreme gradient boosting, multilayer perceptron neural network, and lasso) were applied to clinical and image data from baseline and weeks 4, 8, and 12 for predicting 4 VA outcomes (≥ 15-letter VA gain, ≥ 15-letter VA loss, VA change from baseline, and actual VA) at 2 years. The CATT data from 1029 participants were randomly split for training (n = 717), from which the models were trained using 10-fold cross-validation, and for final validation on a test data set (n = 312).
Performances of ML models were assessed by R and mean absolute error (MAE) for predicting VA change from baseline and actual VA at 2 years, by the area under the receiver operating characteristic curve (AUC) for predicting ≥ 15-letter VA gain and loss from baseline.
Using training data up to week 12, the ML models from cross-validation achieved mean R of 0.24 to 0.29 (MAE = 9.1-9.8 letters) for predicting VA change and 0.37 to 0.41 (MAE = 9.3-10.2 letters) for predicting actual VA at 2 years. The mean AUCs for predicting ≥ 15-letter VA gain and loss at 2 years was 0.84 to 0.85 and 0.58 to 0.73, respectively. In final validation on the test data set up to week 12, the models had an R of 0.33 to 0.38 (MAE = 8.9-9.9 letters) for predicting VA change, an R of 0.37 to 0.45 (MAE = 8.8-10.2 letters) for predicting actual VA at 2 years, and AUCs of 0.85 to 0.87 and 0.67 to 0.79 for predicting ≥ 15-letter VA gain and loss, respectively.
Machine learning models have the potential to predict 2-year VA response to anti-VEGF treatment using clinical and imaging features from the loading dose phase, which can aid in decision-making around treatment protocols for patients with nAMD.
FINANCIAL DISCLOSURE(S): The author(s) have no proprietary or commercial interest in any materials discussed in this article.
评估多种机器学习 (ML) 模型在抗血管内皮生长因子 (anti-VEGF) 治疗湿性年龄相关性黄斑变性 (AMD) 的比较年龄相关性黄斑变性治疗试验 (CATT) 中预测 2 年视力 (VA) 反应的能力,患者患有新生血管性 AMD (nAMD)。
随机临床试验的二次分析。
共有 1029 名 CATT 参与者在研究眼完成了 2 年的随访,随访期间未经治疗的活动性 nAMD 和基线 VA 在 20/25 至 20/320 之间。
应用 5 种 ML 模型(支持向量机、随机森林、极端梯度增强、多层感知机神经网络和套索)对基线和第 4、8 和 12 周的临床和图像数据进行分析,以预测 4 个 VA 结果(≥ 15 字母 VA 增益、≥ 15 字母 VA 损失、VA 基线变化和实际 VA)在 2 年内。1029 名参与者的数据随机分为训练集 (n=717),使用 10 折交叉验证对模型进行训练,并在测试数据集 (n=312) 上进行最终验证。
通过 R 和平均绝对误差 (MAE) 评估 ML 模型在预测 2 年 VA 变化和实际 VA 方面的性能,通过接收者操作特征曲线 (AUC) 评估预测基线时 ≥ 15 字母 VA 增益和损失的能力。
使用训练数据直至第 12 周,交叉验证的 ML 模型对 2 年内 VA 变化的平均 R 为 0.24 至 0.29(MAE=9.1-9.8 个字母),对实际 VA 的平均 R 为 0.37 至 0.41(MAE=9.3-10.2 个字母)。预测 2 年内≥15 个字母 VA 增益和损失的平均 AUC 分别为 0.84 至 0.85 和 0.58 至 0.73。在第 12 周的测试数据集上进行最终验证时,模型对 VA 变化的 R 为 0.33 至 0.38(MAE=8.9-9.9 个字母),对实际 VA 的 R 为 0.37 至 0.45(MAE=8.8-10.2 个字母),预测≥15 个字母 VA 增益和损失的 AUC 分别为 0.85 至 0.87 和 0.67 至 0.79。
机器学习模型具有使用加载剂量阶段的临床和影像学特征预测抗 VEGF 治疗 2 年 VA 反应的潜力,这有助于为 nAMD 患者制定治疗方案。
作者没有与本文讨论的材料有关的专有或商业利益。