Dong Vincent, Sevgi Duriye Damla, Kar Sudeshna Sil, Srivastava Sunil K, Ehlers Justis P, Madabhushi Anant
The Center for Computational Imaging and Personalized Diagnostics, Case Western Reserve University, Department of Biomedical Engineering, Cleveland, OH, United States.
The Tony and Leona Campane Center for Excellence in Image-Guided Surgery and Advanced Imaging Research, Cleveland Clinic, Cleveland, OH, United States.
Front Ophthalmol (Lausanne). 2022;2. doi: 10.3389/fopht.2022.852107. Epub 2022 Aug 12.
Deep learning (DL) is a technique explored within ophthalmology that requires large datasets to distinguish feature representations with high diagnostic performance. There is a need for developing DL approaches to predict therapeutic response, but completed clinical trial datasets are limited in size. Predicting treatment response is more complex than disease diagnosis, where hallmarks of treatment response are subtle. This study seeks to understand the utility of DL for clinical problems in ophthalmology such as predicting treatment response and where large sample sizes for model training are not available.
Four DL architectures were trained using cross-validated transfer learning to classify ultra-widefield angiograms (UWFA) and fluid-compartmentalized optical coherence tomography (OCT) images from a completed clinical trial (PERMEATE) dataset (n=29) as tolerating or requiring extended interval Anti-VEGF dosing. UWFA images (n=217) from the Anti-VEGF study were divided into five increasingly larger subsets to evaluate the influence of dataset size on performance. Class activation maps (CAMs) were generated to identify regions of model attention.
The best performing DL model had a mean AUC of 0.507 ± 0.042 on UWFA images, and highest observed AUC of 0.503 for fluid-compartmentalized OCT images. DL had a best performing AUC of 0.634 when dataset size was incrementally increased. Resulting CAMs show inconsistent regions of interest.
This study demonstrated the limitations of DL for predicting therapeutic response when large datasets were not available for model training. Our findings suggest the need for hand-crafted approaches for complex and data scarce prediction problems in ophthalmology.
深度学习（DL）是眼科领域探索的一种技术，需要大型数据集来区分具有高诊断性能的特征表示。开发用于预测治疗反应的深度学习方法很有必要，但完整的临床试验数据集规模有限。预测治疗反应比疾病诊断更复杂，治疗反应的特征很细微。本研究旨在了解深度学习在眼科临床问题中的效用，例如预测治疗反应，以及在无法获得用于模型训练的大样本量的情况下。
使用交叉验证的迁移学习训练四种深度学习架构，将来自一项已完成的临床试验（PERMEATE）数据集（n = 29）的超广角血管造影（UWFA）和液腔光学相干断层扫描（OCT）图像分类为耐受或需要延长抗血管内皮生长因子（Anti-VEGF）给药间隔。将抗血管内皮生长因子研究中的UWFA图像（n = 217）分为五个越来越大的子集，以评估数据集大小对性能的影响。生成类激活映射（CAM）以识别模型关注区域。
表现最佳的深度学习模型在UWFA图像上的平均曲线下面积（AUC）为0.507±0.042，在液腔OCT图像上观察到的最高AUC为0.503。当数据集大小逐渐增加时，深度学习的最佳AUC为0.634。生成的CAM显示出不一致的感兴趣区域。
本研究证明了在无法获得用于模型训练的大型数据集时，深度学习在预测治疗反应方面的局限性。我们的研究结果表明，对于眼科中复杂且数据稀缺的预测问题，需要采用手工制作的方法。
